W3C

CSS Flexible Box Layout Module Level 1

W3C Last Call Working Draft, 25 September 2014

This version:
http://www.w3.org/TR/2014/WD-css-flexbox-1-20140925/
Latest version:
http://www.w3.org/TR/css-flexbox-1/
Editor’s Draft:
http://dev.w3.org/csswg/css-flexbox/
Previous Versions:
http://www.w3.org/TR/2014/WD-css-flexbox-1-20140325/
http://www.w3.org/TR/2012/CR-css3-flexbox-20120918/
http://www.w3.org/TR/2012/WD-css3-flexbox-20120612/
http://www.w3.org/TR/2012/WD-css3-flexbox-20120322/
http://www.w3.org/TR/2011/WD-css3-flexbox-20111129/
http://www.w3.org/TR/2011/WD-css3-flexbox-20110322/
http://www.w3.org/TR/2009/WD-css3-flexbox-20090723/
Feedback:
www-style@w3.org with subject line “[css-flexbox] … message topic …”(archives)
Test Suite:
http://test.csswg.org/suites/css-flexbox-1_dev/nightly-unstable/
Editors:
Tab Atkins Jr. (Google)
fantasai (Mozilla)
(Microsoft)
Former Editors:
(Microsoft Corporation)
(Mozilla Corporation)
(Mozilla Corporation)
(formerly of Opera Software)
(formerly of Netscape Corporation)
Issues List:
http://dev.w3.org/csswg/css-flexbox-1/issues-lc-20140925

Abstract

The specification describes a CSS box model optimized for user interface design. In the flex layout model, the children of a flex container can be laid out in any direction, and can “flex” their sizes, either growing to fill unused space or shrinking to avoid overflowing the parent. Both horizontal and vertical alignment of the children can be easily manipulated. Nesting of these boxes (horizontal inside vertical, or vertical inside horizontal) can be used to build layouts in two dimensions.

CSS is a language for describing the rendering of structured documents (such as HTML and XML) on screen, on paper, in speech, etc.

Status of this document

This section describes the status of this document at the time of its publication. Other documents may supersede this document. A list of current W3C publications and the latest revision of this technical report can be found in the W3C technical reports index at http://www.w3.org/TR/.

Publication as a Last Call Working Draft does not imply endorsement by the W3C Membership. This is a draft document and may be updated, replaced or obsoleted by other documents at any time. It is inappropriate to cite this document as other than work in progress.

The (archived) public mailing list www-style@w3.org (see instructions) is preferred for discussion of this specification. When sending e-mail, please put the text “css-flexbox” in the subject, preferably like this: “[css-flexbox] …summary of comment…

This document was produced by the CSS Working Group (part of the Style Activity).

This document was produced by a group operating under the 5 February 2004 W3C Patent Policy. W3C maintains a public list of any patent disclosures made in connection with the deliverables of the group; that page also includes instructions for disclosing a patent. An individual who has actual knowledge of a patent which the individual believes contains Essential Claim(s) must disclose the information in accordance with section 6 of the W3C Patent Policy.

This document is governed by the 14 October 2005 W3C Process Document.

This specification is a Last Call Working Draft. All persons are encouraged to review this document and send comments to the www-style mailing list as described above. The deadline for comments is 25 October 2014.

Table of Contents

1 Introduction

This section is not normative.

CSS 2.1 defined four layout modes — algorithms which determine the size and position of boxes based on their relationships with their sibling and ancestor boxes:

This module introduces a new layout mode, flex layout, which is designed for laying out more complex applications and webpages.

1.1 Overview

This section is not normative.

Flex layout is superficially similar to block layout. It lacks many of the more complex text- or document-centric properties that can be used in block layout, such as floats and columns. In return it gains simple and powerful tools for distributing space and aligning content in ways that webapps and complex web pages often need. The contents of a flex container:

Here’s an example of a catalog where each item has a title, an photo, a description, and a purchase button. The designer’s intention is that each entry has the same overall size, that the photo be above the text, and that the purchase buttons aligned at the bottom, regardless of the length of the item’s description. Flex layout makes many aspects of this design easy:

<style>
#deals {
  display: flex;        /* Flex layout so items have equal height  */
  flex-flow: row wrap;  /* Allow items to wrap into multiple lines */
}
.sale-item {
  display: flex;        /* Lay out each item using flex layout */
  flex-flow: column;    /* Lay out item’s contents vertically  */
}
.sale-item > img {
  order: -1;            /* Shift image before other content (in visual order) */
  align-self: center;   /* Center the image cross-wise (horizontally)         */
}
.sale-item > button {
  margin-top: auto;     /* Auto top margin pushes button to bottom */
}
</style>
<section id='deals'>
  <section class='sale-item'>
    <h1>Computer Starter Kit</h1>
    <p>This is the best computer money can buy, if you don’t have much money.
    <ul>
      <li>Computer
      <li>Monitor
      <li>Keyboard
      <li>Mouse
    </ul>
    <img src='images/computer.jpg'
         alt='You get: a white computer with matching peripherals.'>
    <button>BUY NOW</button>
  </section>
  <section class='sale-item'>
    …
  </section>
  …
</section>
You get: a white computer with matching keyboard and monitor.

Computer Starter Kit

This is the best computer money can buy, if you don’t have much money.

  • Computer
  • Monitor
  • Keyboard
  • Mouse
You get: beautiful ASCII art.

Printer

Only capable of printing ASCII art.

  • Paper and ink not included.
An example rendering of the code above.

1.2 Module interactions

This module extends the definition of the display property [CSS21], adding a new block-level and new inline-level display type, and defining a new type of formatting context along with properties to control its layout. None of the properties defined in this module apply to the ::first-line or ::first-letter pseudo-elements.

1.3 Values

This specification follows the CSS property definition conventions from [CSS21]. Value types not defined in this specification are defined in CSS Level 2 Revision 1 [CSS21]. Other CSS modules may expand the definitions of these value types: for example [CSS3VAL], when combined with this module, expands the definition of the <length> value type as used in this specification.

In addition to the property-specific values listed in their definitions, all properties defined in this specification also accept the inherit keyword as their property value. For readability it has not been repeated explicitly.

2 Flex Layout Box Model and Terminology

A flex container is the box generated by an element with a computed display of flex or inline-flex. In-flow children of a flex container are called flex items and are laid out using the flex layout model.

Unlike block and inline layout, whose layout calculations are biased to the block and inline flow directions, flex layout is biased to the flex directions. To make it easier to talk about flex layout, this section defines a set of flex flow–relative terms. The flex-flow value determines how these terms map to physical directions (top/right/bottom/left), axes (vertical/horizontal), and sizes (width/height).

An illustration of the various directions and sizing terms as applied to a row flex container.
main axis
main dimension
The main axis of a flex container is the primary axis along which flex items are laid out. It extends in the main dimension.
main-start
main-end
The flex items are placed within the container starting on the main-start side and going toward the main-end side.
main size
main size property
A flex item’s width or height, whichever is in the main dimension, is the item’s main size. The flex item’s main size property is either the width or height property, whichever is in the main dimension.
cross axis
cross dimension
The axis perpendicular to the main axis is called the cross axis. It extends in the cross dimension.
cross-start
cross-end
Flex lines are filled with items and placed into the container starting on the cross-start side of the flex container and going toward the cross-end side.
cross size
cross size property
The width or height of a flex item, whichever is in the cross dimension, is the item’s cross size. The cross size property is whichever of width or height that is in the cross dimension.

3 Flex Containers: the flex and inline-flex display values

Name:display
New values:flex | inline-flex
flex
This value causes an element to generate a block-level flex container box.
inline-flex
This value causes an element to generate an inline-level flex container box.

A flex container establishes a new flex formatting context for its contents. This is the same as establishing a block formatting context, except that flex layout is used instead of block layout: floats do not intrude into the flex container, and the flex container’s margins do not collapse with the margins of its contents. Flex containers form a containing block for their contents exactly like block containers do. [CSS21] The overflow property applies to flex containers.

Flex containers are not block containers, and so some properties that were designed with the assumption of block layout don’t apply in the context of flex layout. In particular:

If an element’s specified display is inline-flex, then its display property computes to flex in certain circumstances: the table in CSS 2.1 Section 9.7 is amended to contain an additional row, with inline-flex in the "Specified Value" column and flex in the "Computed Value" column.

4 Flex Items

Loosely speaking, the flex items of a flex container are the in-flow boxes generated from the contents of the element that generated the flex container.

Each in-flow child of a flex container becomes a flex item, and each contiguous run of text that is directly contained inside a flex container is wrapped in an anonymous flex item. However, an anonymous flex item that contains only white space (i.e. characters that can be affected by the white-space property) is not rendered, as if it were display:none.

A flex item establishes a new formatting context for its contents. The type of this formatting context is determined by its display value, as usual. However, flex items themselves are flex-level boxes, not block-level boxes: they participate in their container’s flex formatting context, not in a block formatting context.

Examples of flex items:

<div style="display:flex">

    <!-- flex item: block child -->
    <div id="item1">block</div>

    <!-- flex item: floated element; floating is ignored -->
    <div id="item2" style="float: left;">float</div>

    <!-- flex item: anonymous block box around inline content -->
    anonymous item 3

    <!-- flex item: inline child -->
    <span>
        item 4
        <!-- flex items do not split around blocks -->
        <div id=not-an-item>item 4</div>
        item 4
    </span>
</div>

Authors reading this spec may want to skip past the following box-generation details.

If the specified display-outside of an in-flow child of an element that generates a flex container is inline-level, it computes to block-level. (This effectively converts any inline display values to their block equivalents.)

Note: Some values of display-outside trigger the creation of anonymous boxes around the original box. It’s the outermost box, the direct child of the flex container box, that becomes a flex item. For example, given two contiguous child elements with display:table-cell, an anonymous table wrapper box around them becomes the flex item.

On a flex item with display: table, the table wrapper box becomes the flex item, and the order and align-self properties apply to it. The contents of any caption boxes contribute to the calculation of the table wrapper box’s min-content and max-content sizes. However, like width and height, the flex longhands apply to the table box as follows: the flex item’s final size is calculated by performing layout as if the distance between the table wrapper box’s edges and the table box’s content edges were all part of the table box’s border+padding area, and the table box were the flex item.

4.1 Absolutely-Positioned Flex Children

An absolutely-positioned child of a flex container does not participate in flex layout. However, it does participate in the reordering step (see order), which has an effect in their painting order.

The static position of an absolutely-positioned child of a flex container is determined such that the child is positioned as if it were the sole flex item in the flex container, assuming both the child and the flex container were fixed-size boxes of their used size.


In other words, the static position of an absolutely positioned child of a flex container is determined after flex layout by setting the child’s static-position rectangle to the flex container’s content box, then aligning the absolutely positioned child within this rectangle according to the justify-content value of the flex container and the align-self value of the child itself.

The effect of this is that if you set, for example, align-content: center; on an absolutely-positioned child of a flex container, the child’s static position will center it in the flex container’s cross axis.
The static-position rectangle is the alignment container used to determine the static-position offsets of an absolutely-positioned box. In block layout it corresponds to the position of the “hypothetical box” described in CSS2.1§10.3.7. (Since it has no alignment properties, CSS2.1 always uses a block-start inline-start alignment of the absolutely-positioned box within the static-position rectangle.) Note that this definition will eventually move to the CSS Positioning module.

4.2 Flex Item Margins and Paddings

The margins of adjacent flex items do not collapse. Auto margins absorb extra space in the corresponding dimension and can be used for alignment and to push adjacent flex items apart; see Aligning with auto margins.

Percentage margins and paddings on flex items are always resolved against their respective dimensions; unlike blocks, they do not always resolve against the inline dimension of their containing block.

4.3 Flex Item Z-Ordering

Flex items paint exactly the same as inline blocks [CSS21], except that order-modified document order is used in place of raw document order, and z-index values other than auto create a stacking context even if position is static.

Note: Descendants that are positioned outside a flex item still participate in any stacking context established by the flex item.

4.4 Collapsed Items

Specifying visibility:collapse on a flex item causes it to become a collapsed flex item, producing an effect similar to visibility:collapse on a table-row or table-column: the collapsed flex item is removed from rendering entirely, but leaves behind a "strut" that keeps the flex line’s cross-size stable. Thus, if a flex container has only one flex line, dynamically collapsing or uncollapsing items is guaranteed to have no effect on the flex container’s cross size and won’t cause the rest of the page’s layout to "wobble". Flex line wrapping is re-done after collapsing, however, so the cross-size of a flex container with multiple lines might or might not change.

Though collapsed flex items aren’t rendered, they do appear in the formatting structure. Therefore, unlike on display:none items [CSS21], effects that depend on a box appearing in the formatting structure (like incrementing counters or running animations and transitions) still operate on collapsed items.

In the following example, a sidebar is sized to fit its content. visibility: collapse is used to dynamically hide parts of a navigation sidebar without affecting its width, even though the widest item (“Architecture”) is in a collapsed section.

Sample rendering for example code below
<style>
  @media (min-width: 60em) {
    /* two column layout only when enough room (relative to default text size) */
    header + div { display: flex; }
    #main {
      flex: 1;         /* Main takes up all remaining space */
      order: 1;        /* Place it after (to the right of) the navigation */
      min-width: 12em; /* Optimize main content area sizing */
    }
  }
  /* menu items use flex layout so that visibility:collapse will work */
  nav > ul > li {
    display: flex;
    flex-flow: column;
  }
  /* dynamically collapse submenus when not targetted */
  nav > ul > li:not(:target):not(:hover) > ul {
    visibility: collapse;
  }
</style>
…
</header>
<div>
  <article id="main">
    Interesting Stuff to Read
  </article>
  <nav>
    <ul>
      <li id="nav-about"><a href="#nav-about">About</a>
        …
      <li id="nav-projects"><a href="#nav-projects">Projects</a>
        <ul>
          <li><a href="…">Art</a>
          <li><a href="…">Architecture</a>
          <li><a href="…">Music</a>
        </ul>
      <li id="nav-interact"><a href="#nav-interact">Interact</a>
        …
    </ul>
  </nav>
</div>
<footer>
…

To compute the size of the strut, flex layout is first performed with all items uncollapsed, and then re-run with each collapsed item replaced by a strut that maintains the original cross-size of the item’s original line. See the Flex Layout Algorithm for the normative definition of how visibility:collapse interacts with flex layout.

Note that using visibility:collapse on any flex items will cause the flex layout algorithm to repeat partway through, re-running the most expensive steps. It’s recommended that authors continue to use display:none to hide items if the items will not be dynamically collapsed and uncollapsed, as that is more efficient for the layout engine. (Since only part of the steps need to be repeated when visibility is changed, however, 'visibility: collapse' is still recommended for dynamic cases.)

4.5 Implied Minimum Size of Flex Items

To provide a more reasonable default minimum size for flex items, this specification introduces a new auto value as the initial value of the min-width and min-height properties defined in CSS 2.1. [CSS21]

Name:min-width, min-height
New values:auto
New computed value:the percentage as specified or the absolute length or a keyword
New initial value:auto
auto
On a flex item whose overflow is visible in the main axis, when specified on the flex item’s main-axis min-size property, the following table gives the minimum size:
Clamped Size Transferred Size Minimum Size
content size
min(clamped size, content size)
min(transferred size, content size)
min(clamped size, content size)

Where:

clamped size
is defined if the item’s flex-basis is main-size and its computed main size property is definite, and is that size (clamped by its max main size property if it’s definite).
transferred size
is defined if the item has an intrinsic aspect ratio and its computed cross size property is definite, and is that size (clamped by its min and max cross size properties if they are definite), converted through the aspect ratio.
content size
is the min-content size in the main axis, clamped, if it has an aspect ratio, by any definite min and max cross size properties converted through the aspect ratio, and then further clamped by the max main size property if that is definite.

Otherwise, this keyword computes to 0 (unless otherwise defined by a future specification).

Note that while a content-based minimum size is often appropriate, and helps prevent content from overlapping or spilling outside its container, in some cases it is not:

In particular, if flex sizing is being used for a major content area of a document, it is better to set an explicit font-relative minimum width such as min-width: 12em. A content-based minimum width could result in a large table or large image stretching the size of the entire content area into an overflow zone, and thereby making lines of text gratuitously long and hard to read.

Note also, when content-based sizing is used on an item with large amounts of content, the layout engine must traverse all of this content before finding its minimum size, whereas if the author sets an explicit minimum, this is not necessary. (For items with small amounts of content, however, this traversal is trivial and therefore not a performance concern.)

In order to prevent cycling sizing, the auto value of min-height and max-height does not factor into the percentage size resolution of the box’s contents. For example, a percentage-height block whose flex item parent has height: 120em; min-height: auto will size itself against height: 120em regardless of the impact that min-height might have on the used size of the flex item.

5 Ordering and Orientation

The contents of a flex container can be laid out in any direction and in any order. This allows an author to trivially achieve effects that would previously have required complex or fragile methods, such as hacks using the float and clear properties. This functionality is exposed through the flex-direction, flex-wrap, and order properties.

The reordering capabilities of flex layout intentionally affect only the visual rendering, leaving speech order and navigation based on the source order. This allows authors to manipulate the visual presentation while leaving the source order intact for non-CSS UAs and for linear models such as speech and sequential navigation. See Reordering and Accessibility and the Flex Layout Overview for examples that use this dichotomy to improve accessibility.

Authors must not use order or the *-reverse values of flex-flow/flex-direction as a substitute for correct source ordering, as that can ruin the accessibility of the document.

5.1 Flex Flow Direction: the flex-direction property

Name:flex-direction
Value:row | row-reverse | column | column-reverse
Initial:row
Applies to:flex containers
Inherited:no
Media:visual
Computed value:specified value

The flex-direction property specifies how flex items are placed in the flex container, by setting the direction of the flex container’s main axis. This determines the direction that flex items are laid out in.

row
The flex container’s main axis has the same orientation as the inline axis of the current writing mode. The main-start and main-end directions are equivalent to the inline-start and inline-end directions, respectively, of the current writing mode.
row-reverse
Same as row, except the main-start and main-end directions are swapped.
column
The flex container’s main axis has the same orientation as the block axis of the current writing mode. The main-start and main-end directions are equivalent to the block-start and block-end directions, respectively, of the current writing mode.
column-reverse
Same as column, except the main-start and main-end directions are swapped.

The reverse values do not reverse box ordering: like writing-mode and direction [CSS3-WRITING-MODES], they only change the direction of flow. Painting order, speech order, and sequential navigation orders are not affected.

5.2 Flex Line Wrapping: the flex-wrap property

Name:flex-wrap
Value:nowrap | wrap | wrap-reverse
Initial:nowrap
Applies to:flex containers
Inherited:no
Media:visual
Computed value:specified value

The flex-wrap property controls whether the flex container is single-line or multi-line, and the direction of the cross-axis, which determines the direction new lines are stacked in.

nowrap
The flex container is single-line. The cross-start direction is equivalent to either the inline-start or block-start direction of the current writing mode, whichever is in the cross axis, and the cross-end direction is the opposite direction of cross-start.
wrap
The flex container is multi-line. The cross-start direction is equivalent to either the inline-start or block-start direction of the current writing mode, whichever is in the cross axis, and the cross-end direction is the opposite direction of cross-start.
wrap-reverse
Same as wrap, except the cross-start and cross-end directions are swapped.

5.3 Flex Direction and Wrap: the flex-flow shorthand

Name:flex-flow
Value:<flex-direction> || <flex-wrap>
Initial:see individual properties
Applies to:flex containers
Inherited:see individual properties
Media:visual
Computed value:see individual properties

The flex-flow property is a shorthand for setting the flex-direction and flex-wrap properties, which together define the flex container’s main and cross axes.

Some examples of valid flows in an English (left-to-right, horizontal writing mode) document:

div { flex-flow: row; }
/* Initial value. Main-axis is
   inline, no wrap. */
div { flex-flow: column wrap; }
/* Main-axis is block-direction (top to bottom)
   and lines wrap in the inline direction (rightwards). */
div { flex-flow: row-reverse wrap-reverse; }
/* Main-axis is the opposite of inline direction
   (right to left). New lines wrap upwards. */

Note that the flex-flow directions are writing mode sensitive. In vertical Japanese, for example, a row flex container lays out its contents from top to bottom, as seen in this example:

English Japanese
flex-flow: row wrap;
writing-mode: horizontal-tb;
flex-flow: row wrap;
writing-mode: vertical-rl;

5.4 Display Order: the order property

Flex items are, by default, displayed and laid out in the same order as they appear in the source document. The order property can be used to change this ordering.

Name:order
Value:<integer>
Initial:0
Applies to:flex items and absolutely-positioned children of flex containers
Inherited:no
Media:visual
Computed value:specified value
Animatable:yes

The order property controls the order in which children of a flex container appear within the flex container, by assigning them to ordinal groups. It takes a single <integer> value, which specifies which ordinal group the flex item belongs to.

A flex container lays out its content in order-modified document order, starting from the lowest numbered ordinal group and going up. Items with the same ordinal group are laid out in the order they appear in the source document. This also affects the painting order [CSS21], exactly as if the flex items were reordered in the source document.

The following figure shows a simple tabbed interface, where the tab for the active pane is always first:

This could be implemented with the following CSS (showing only the relevant code):

.tabs {
  display: flex;
}
.tabs > * {
  min-width: min-content;
  /* Prevent tabs from getting too small for their content. */
}
.tabs > .current {
  order: -1; /* Lower than the default of 0 */
}

Unless otherwise specified by a future specification, this property has no effect on boxes that are not children of a flex container.

5.4.1 Reordering and Accessibility

The order property does not affect ordering in non-visual media (such as speech). Likewise, order does not affect the default traversal order of sequential navigation modes (such as cycling through links, see e.g. nav-index [CSS3UI] or tabindex [HTML40]). Authors must use order only for visual, not logical, reordering of content; style sheets that use order to perform logical reordering are non-conforming.

This is so that non-visual media and non-CSS UAs, which typically present content linearly, can rely on a logical source order, while order is used to tailor the visual order. (Since visual perception is two-dimensional and non-linear, the desired visual order is not always logical.)

Many web pages have a similar shape in the markup, with a header on top, a footer on bottom, and then a content area and one or two additional columns in the middle. Generally, it’s desirable that the content come first in the page’s source code, before the additional columns. However, this makes many common designs, such as simply having the additional columns on the left and the content area on the right, difficult to achieve. This has been addressed in many ways over the years, often going by the name "Holy Grail Layout" when there are two additional columns. order makes this trivial. For example, take the following sketch of a page’s code and desired layout:

<!DOCTYPE html>
<header>...</header>
<div id='main'>
   <article>...</article>
   <nav>...</nav>
   <aside>...</aside>
</div>
<footer>...</footer>
In this page the header is at the top and the footer at the bottom, but the article is in the center, flanked by the nav on the right and the aside on the left.

This layout can be easily achieved with flex layout:

#main { display: flex; }
#main > article { order: 2; min-width: 12em; flex:1; }
#main > nav     { order: 1; width: 200px; }
#main > aside   { order: 3; width: 200px; }

As an added bonus, the columns will all be equal-height by default, and the main content will be as wide as necessary to fill the screen. Additionally, this can then be combined with media queries to switch to an all-vertical layout on narrow screens:

@media all and (max-width: 600px) {
  /* Too narrow to support three columns */
  #main { flex-flow: column; }
  #main > article, #main > nav, #main > aside {
    /* Return them to document order */
    order: 0; width: auto;
  }
}

(Further use of multi-line flex containers to achieve even more intelligent wrapping left as an exercise for the reader.)

6 Flex Lines

Flex items in a flex container are laid out and aligned within flex lines, hypothetical containers used for grouping and alignment by the layout algorithm. A flex container can be either single-line or multi-line, depending on the flex-wrap property:

This example shows four buttons that do not fit horizontally.

<style>
#flex {
  display: flex;
  flex-flow: row wrap;
  width: 300px;
}
.item {
  width: 80px;
}
</style>

<div id="flex">
  <div class='item'>1</div>
  <div class='item'>2</div>
  <div class='item'>3</div>
  <div class='item'>4</div>
</div>

Since the container is 300px wide, only three of the items fit onto a single line. They take up 240px, with 60px left over of remaining space. Because the flex-flow property specifies a multi-line flex container (due to the wrap keyword appearing in its value), the flex container will create an additional line to contain the last item.

An example rendering of the multi-line flex container.

Once content is broken into lines, each line is laid out independently; flexible lengths and the justify-content and align-self properties only consider the items on a single line at a time.

When a flex container has multiple lines, the cross size of each line is the minimum size necessary to contain the flex items on the line (after aligment due to align-self), and the lines are aligned within the flex container with the align-content property. When a flex container (even a multi-line one) has only one line, the cross size of the line is the cross size of the flex container, and align-content has no effect. The main size of a line is always the same as the main size of the flex container’s content box.

Here’s the same example as the previous, except that the flex items have all been given flex: auto. The first line has 60px of remaining space, and all of the items have the same flexibility, so each of the three items on that line will receives 20px of extra width, ending up 100px wide. The remaining item is on a line of its own and will stretch to the entire width of the line, or 300px.

A rendering of the same as above, but with the items all given flex: auto.

7 Flexibility

The defining aspect of flex layout is the ability to make the flex items “flex”, altering their width/height to fill the available space in the main dimension. This is done with the flex property. A flex container distributes free space to its items proportional to their flex grow factor, or shrinks them to prevent overflow proportional to their flex shrink factor.

7.1 The flex Shorthand

Name:flex
Value:none | auto | [ <‘flex-grow’> <‘flex-shrink’>? || <‘flex-basis’> ]
Initial:see individual properties
Applies to:flex items
Inherited:see individual properties
Media:visual
Computed value:see individual properties
Canonical order:per grammar

The flex property specifies the components of a flexible length: the flex grow factor and flex shrink factor, and the flex basis. When a box is a flex item, flex is consulted instead of the main size property to determine the main size of the box. If a box is not a flex item, flex has no effect.

<‘flex-grow’>

This <number> component sets flex-grow longhand and specifies the flex grow factor, which determines how much the flex item will grow relative to the rest of the flex items in the flex container when positive free space is distributed. When omitted, it is set to 1.

<‘flex-shrink’>

This <number> component sets flex-shrink longhand and specifies the flex shrink factor, which determines how much the flex item will shrink relative to the rest of the flex items in the flex container when negative free space is distributed. When omitted, it is set to 1. The flex shrink factor is multiplied by the flex basis when distributing negative space.

<‘flex-basis’>

This component, which takes the same values as the width property, sets the flex-basis longhand and specifies the flex basis: the initial main size of the flex item, before free space is distributed according to the flex factors. When omitted from the flex shorthand, its specified value is 0%.

If the specified flex-basis is main-size, the used flex basis is the computed value of the flex item’s main size property.

A diagram showing the difference between "absolute" flex (starting from a basis of zero) and "relative" flex (starting from a basis of the item’s content size). The three items have flex factors of 1, 1, and 2, respectively.
auto
The keyword auto expands to 1 1 main-size.
none

The keyword none expands to 0 0 main-size.

The initial values of the flex components are equivalent to flex: 0 1 main-size.

Note that the initial values of flex-grow and flex-basis are different from their defaults when omitted in the flex shorthand. This so that the flex shorthand can better accommodate the most common cases.

A unitless zero that is not already preceded by two flex factors must be interpreted as a flex factor. To avoid misinterpretation or invalid declarations, authors must specify a zero <‘flex-basis’> component with a unit or precede it by two flex factors.

7.2 Common Values of flex

This section is informative.

The list below summarizes the effects of the most common flex values:

flex: 0 main-size
flex: initial
Equivalent to flex: 0 1 main-size. (This is the initial value.) Sizes the item based on the width/height properties. (If the item’s main size property computes to auto, this will size the flex item based on its contents.) Makes the flex item inflexible when there is positive free space, but allows it to shrink to its min-size when there is insufficient space. The alignment abilities or auto margins can be used to align flex items along the main axis.
flex: auto
Equivalent to flex: 1 1 main-size. Sizes the item based on the width/height properties, but makes them fully flexible, so that they absorb any free space along the main axis. If all items are either flex: auto, flex: initial, or flex: none, any positive free space after the items have been sized will be distributed evenly to the items with flex: auto.
flex: none
Equivalent to flex: 0 0 main-size. This value sizes the item according to the width/height properties, but makes the flex item fully inflexible. This is similar to initial, except that flex items are not allowed to shrink, even in overflow situations.
flex: <positive-number>
Equivalent to flex: <positive-number> 1 0%. Makes the flex item flexible and sets the flex basis to zero, resulting in an item that receives the specified proportion of the free space in the flex container. If all items in the flex container use this pattern, their sizes will be proportional to the specified flex factor.

By default, flex items won’t shrink below their minimum content size (the length of the longest word or fixed-size element). To change this, set the min-width or min-height property. (See Implied Minimum Size of Flex Items.)

7.3 Components of Flexibility

Individual components of flexibility can be controlled by independent longhand properties.

Note: Authors are encouraged to control flexibility using the flex shorthand rather than with component properties, as the shorthand correctly resets any unspecified components to accommodate common uses.

7.3.1 The flex-grow property

Name:flex-grow
Value:<number>
Initial:0
Applies to:flex items
Inherited:no
Media:visual
Computed value:specified value
Animatable:yes

Note: Authors are encouraged to control flexibility using the flex shorthand rather than with component properties, as the shorthand correctly resets any unspecified components to accommodate common uses.

The flex-grow property sets the flex grow factor to the provided <number>. Negative numbers are invalid.

7.3.2 The flex-shrink property

Name:flex-shrink
Value:<number>
Initial:1
Applies to:flex items
Inherited:no
Media:visual
Computed value:specified value
Animatable:yes

Note: Authors are encouraged to control flexibility using the flex shorthand rather than with component properties, as the shorthand correctly resets any unspecified components to accommodate common uses.

The flex-shrink property sets the flex shrink factor to the provided <number>. Negative numbers are invalid.

7.3.3 The flex-basis property

Name:flex-basis
Value:main-size | <‘width’>
Initial:main-size
Applies to:flex items
Inherited:no
Media:visual
Computed value:as specified, with lengths made absolute
Percentages:relative to the flex container’s inner main size
Animatable:as width

Note: Authors are encouraged to control flexibility using the flex shorthand rather than with component properties, as the shorthand correctly resets any unspecified components to accommodate common uses.

The flex-basis property sets the flex basis. It accepts the same values as the width and height property.

The main-size keyword, when specified on a flex item, retrieves the value of the main size property.

This value was originally named auto, but that was problematic because it left us with no keyword to represent an automatic flex basis. There were two options for fixing this:
auto for automatic sizing; main-axis for size lookup
The advantage of this is primarily that auto means “automatic” both when it is specified for flex-basis and when it is pulled in from width/height. It has the disadvantage (due to back-compat concerns) that the flex shorthand had to be tweaked to expand flex: auto to flex: auto main-size rather than to flex: 1 auto. It is the currently-implemented option, but may be unshippable due to back-compat constraints.
auto for size lookup; content1 for automatic sizing
The advantage of this is that it will not break existing content: a flex-basis keyword for automatic sizing did not exist in the original CR, so this option adds a new keyword but does not rename existing functionality. It therefore avoids the flex: autoflex: 1 auto inequality.

The CSSWG is waiting for feedback from implementers on this issue.

1 Alternate keyword suggestions welcome.

The flex basis is resolved the same way as width in horizontal writing modes [CSS21]: percentage values of flex-basis are resolved against the flex item’s containing block, i.e. its flex container, and if that containing block’s size is indefinite, the result is the same as a main size of auto. Similarly, flex-basis determines the size of the content box, unless otherwise specified such as by box-sizing [CSS3UI].

8 Alignment

After a flex container’s contents have finished their flexing and the dimensions of all flex items are finalized, they can then be aligned within the flex container.

The margin properties can be used to align items in a manner similar to, but more powerful than, what margins can do in block layout. Flex items also respect the alignment properties from CSS Box Alignment, which allow easy keyword-based alignment of items in both the main axis and cross axis. These properties make many common types of alignment trivial, including some things that were very difficult in CSS 2.1, like horizontal and vertical centering.

While the alignment properties are defined in CSS Box Alignment, Flexible Box Layout reproduces the definitions of the relevant ones here so as to not create a normative dependency that may slow down advancement of the spec. These properties apply only to flex layout until CSS Box Alignment Level 3 is finished and defines their effect for other layout modes.

8.1 Aligning with auto margins

This section is non-normative. The normative definition of how margins affect flex items is in the Flex Layout Algorithm section.

Auto margins on flex items have an effect very similar to auto margins in block flow:

Note that, if free space is distributed to auto margins, the alignment properties will have no effect in that dimension because the margins will have stolen all the free space left over after flexing.

One use of auto margins in the main axis is to separate flex items into distinct "groups". The following example shows how to use this to reproduce a common UI pattern - a single bar of actions with some aligned on the left and others aligned on the right.

Sample rendering of the code below.
<style>
nav > ul {
  display: flex;
}
nav > ul > li {
  min-width: min-content;
  /* Prevent items from getting too small for their content. */
}
nav > ul > #login {
  margin-left: auto;
}
</style>
<nav>
  <ul>
    <li><a href=/about>About</a>
    <li><a href=/projects>Projects</a>
    <li><a href=/interact>Interact</a>
    <li id='login'><a href=/login>Login</a>
  </ul>
</nav>

The figure below illustrates the difference in cross-axis alignment in overflow situations between using auto margins and using the alignment properties.

About
Authoritarianism
Blog
About
Authoritarianism
Blog
The items in the figure on the left are centered with margins, while those in the figure on the right are centered with align-self. If this column flex container was placed against the left edge of the page, the margin behavior would be more desirable, as the long item would be fully readable. In other circumstances, the true centering behavior might be better.

8.2 Axis Alignment: the justify-content property

Name:justify-content
Value:flex-start | flex-end | center | space-between | space-around
Initial:flex-start
Applies to:flex containers
Inherited:no
Media:visual
Computed value:specified value

The justify-content property aligns flex items along the main axis of the current line of the flex container. This is done after any flexible lengths and any auto margins have been resolved. Typically it helps distribute extra free space leftover when either all the flex items on a line are inflexible, or are flexible but have reached their maximum size. It also exerts some control over the alignment of items when they overflow the line.

flex-start
Flex items are packed toward the start of the line. The main-start margin edge of the first flex item on the line is placed flush with the main-start edge of the line, and each subsequent flex item is placed flush with the preceding item.
flex-end
Flex items are packed toward the end of the line. The main-end margin edge of the last flex item is placed flush with the main-end edge of the line, and each preceding flex item is placed flush with the subsequent item.
center
Flex items are packed toward the center of the line. The flex items on the line are placed flush with each other and aligned in the center of the line, with equal amounts of empty space between the main-start edge of the line and the first item on the line and between the main-end edge of the line and the last item on the line. (If the leftover free-space is negative, the flex items will overflow equally in both directions.)
space-between
Flex items are evenly distributed in the line. If the leftover free-space is negative or there is only a single flex item on the line, this value is identical to flex-start. Otherwise, the main-start margin edge of the first flex item on the line is placed flush with the main-start edge of the line, the main-end margin edge of the last flex item on the line is placed flush with the main-end edge of the line, and the remaining flex items on the line are distributed so that the spacing between any two adjacent items is the same.
space-around
Flex items are evenly distributed in the line, with half-size spaces on either end. If the leftover free-space is negative or there is only a single flex item on the line, this value is identical to center. Otherwise, the flex items on the line are distributed such that the spacing between any two adjacent flex items on the line is the same, and the spacing between the first/last flex items and the flex container edges is half the size of the spacing between flex items.

An illustration of the five justify-content keywords and their effects on a flex container with three colored items.

8.3 Cross-axis Alignment: the align-items and align-self properties

Name:align-items
Value:flex-start | flex-end | center | baseline | stretch
Initial:stretch
Applies to:flex containers
Inherited:no
Media:visual
Computed value:specified value
Name:align-self
Value:auto | flex-start | flex-end | center | baseline | stretch
Initial:auto
Applies to:flex items
Inherited:no
Media:visual
Computed value:auto computes to parent’s align-items value; otherwise as specified

Flex items can be aligned in the cross axis of the current line of the flex container, similar to justify-content but in the perpendicular direction. align-items sets the default alignment for all of the flex container’s items, including anonymous flex items. align-self allows this default alignment to be overridden for individual flex items. (For anonymous flex items, align-self always matches the value of align-items on their associated flex container.)

If either of the flex item’s cross-axis margins are auto, align-self has no effect.

A value of auto for align-self computes to the value of align-items on the element’s parent, or stretch if the element has no parent. The alignments are defined as:

flex-start
The cross-start margin edge of the flex item is placed flush with the cross-start edge of the line.
flex-end
The cross-end margin edge of the flex item is placed flush with the cross-end edge of the line.
center
The flex item’s margin box is centered in the cross axis within the line. (If the cross size of the flex line is less than that of the flex item, it will overflow equally in both directions.)
baseline

If the flex item’s inline axis is the same as the cross axis, this value is identical to flex-start.

Otherwise, it participates in baseline alignment: all participating flex items on the line are aligned such that their baselines align, and the item with the largest distance between its baseline and its cross-start margin edge is placed flush against the cross-start edge of the line.

stretch
If the cross size property of the flex item computes to auto, and either of the cross-axis margins are auto, the flex item is stretched. Its used value is the length necessary to make the cross size of the item’s margin box as close to the same size as the line as possible, while still respecting the constraints imposed by min-height/min-width/max-height/max-width.

Note: If the flex container’s height is constrained this value may cause the contents of the flex item to overflow the item.

The cross-start margin edge of the flex item is placed flush with the cross-start edge of the line.

An illustration of the five align-items keywords and their effects on a flex container with four colored items.

8.4 Packing Flex Lines: the align-content property

Name:align-content
Value:flex-start | flex-end | center | space-between | space-around | stretch
Initial:stretch
Applies to:multi-line flex containers
Inherited:no
Media:visual
Computed value:specified value

The align-content property aligns a flex container’s lines within the flex container when there is extra space in the cross-axis, similar to how justify-content aligns individual items within the main-axis. Note, this property has no effect when the flex container has only a single line. Values have the following meanings:

flex-start
Lines are packed toward the start of the flex container. The cross-start edge of the first line in the flex container is placed flush with the cross-start edge of the flex container, and each subsequent line is placed flush with the preceding line.
flex-end
Lines are packed toward the end of the flex container. The cross-end edge of the last line is placed flush with the cross-end edge of the flex container, and each preceding line is placed flush with the subsequent line.
center
Lines are packed toward the center of the flex container. The lines in the flex container are placed flush with each other and aligned in the center of the flex container, with equal amounts of empty space between the cross-start content edge of the flex container and the first line in the flex container, and between the cross-end content edge of the flex container and the last line in the flex container. (If the leftover free-space is negative, the lines will overflow equally in both directions.)
space-between
Lines are evenly distributed in the flex container. If the leftover free-space is negative this value is identical to flex-start. Otherwise, the cross-start edge of the first line in the flex container is placed flush with the cross-start content edge of the flex container, the cross-end edge of the last line in the flex container is placed flush with the cross-end content edge of the flex container, and the remaining lines in the flex container are distributed so that the spacing between any two adjacent lines is the same.
space-around
Lines are evenly distributed in the flex container, with half-size spaces on either end. If the leftover free-space is negative this value is identical to center. Otherwise, the lines in the flex container are distributed such that the spacing between any two adjacent lines is the same, and the spacing between the first/last lines and the flex container edges is half the size of the spacing between flex lines.
stretch
Lines stretch to take up the remaining space. If the leftover free-space is negative, this value is identical to flex-start. Otherwise, the free-space is split equally between all of the lines, increasing their cross size.

Note: Only flex containers with multiple lines ever have free space in the cross-axis for lines to be aligned in, because in a flex container with a single line the sole line automatically stretches to fill the space.

An illustration of the align-content keywords and their effects on a multi-line flex container.

8.5 Flex Baselines

The baselines of a flex container are determined as follows (after reordering with order):

main-axis baseline
  1. If any of the flex items on the flex container’s first line participate in baseline alignment, the flex container’s main-axis baseline is the baseline of those flex items.
  2. Otherwise, if the flex container has at least one flex item, and its first flex item has a baseline parallel to the flex container’s main axis, the flex container’s main-axis baseline is that baseline.
  3. Otherwise, the flex container’s main-axis baseline is synthesized from the first item’s content box, or, failing that, from the flex container’s content box.
cross-axis baseline
  1. If the flex container has at least one flex item, and its first flex item has a baseline parallel to the flex container’s cross axis, the flex container’s cross-axis baseline is that baseline.
  2. Otherwise, the flex container’s cross-axis baseline is synthesized from the first item’s content box, or, failing that, from the flex container’s content box.

When calculating the baseline according to the above rules, if the box contributing a baseline has an overflow value that allows scrolling, the box must be treated as being in its initial scroll position for the purpose of determining its baseline.

When determining the baseline of a table cell, a flex container provides a baseline just as a line box or table-row does. [CSS21]

See Baseline Rules in [CSS3-ALIGN] for more information on baseline alignment.

9 Flex Layout Algorithm

This section contains normative algorithms detailing the exact layout behavior of a flex container and its contents. The algorithms here are written to optimize readability and theoretical simplicity, and may not necessarily be the most efficient. Implementations may use whatever actual algorithms they wish, but must produce the same results as the algorithms described here.

Note: This section is mainly intended for implementors. Authors writing web pages should generally be served well by the individual property descriptions, and do not need to read this section unless they have a deep-seated urge to understand arcane details of CSS layout.

The following sections define the algorithm for laying out a flex container and its contents.

9.1 Initial Setup

  1. Generate anonymous flex items as described in the Flex Items section.
  2. Re-order the flex items (and any absolutely positioned flex container children) according to their order. The flex items with the lowest (most negative) order values are first in the ordering. If multiple flex items share an order value, they’re ordered by document order. This effectively changes the order of their boxes in the box-tree, and how the rest of this algorithm deals with the generated flex items.

    Note: Absolutely-positioned children of a flex container do not participate in flex layout, but are reordered together with any flex item children.

9.2 Line Length Determination

  1. Determine the available main and cross space for the flex items. For each dimension, if that dimension of the flex container’s content box is a definite size, use that; otherwise, subtract the flex container’s margin, border, and padding from the space available to the flex container in that dimension and use that value. This might result in an infinite value.

    For example, the available space to a flex item in a floated auto-sized flex container is:

    • the width of the flex container’s containing block minus the flex container’s margin, border, and padding in the horizontal dimension
    • infinite in the vertical dimension
  2. Determine the flex base size and hypothetical main size of each item:
    1. If the item has a definite used flex basis, that’s the flex base size.
    2. If the flex item has ... then the flex base size is calculated from its inner cross size and the flex item’s intrinsic aspect ratio.
    3. If the used flex basis is auto or depends on its available size, and the flex container is being sized under a min-content or max-content constraint (e.g. when performing automatic table layout [CSS21]), size the item under that constraint. The flex base size is the item’s resulting main size.
    4. Otherwise, if the used flex basis is auto or depends on its available size, the available main size is infinite, and the flex item’s inline axis is parallel to the main axis, lay the item out using the rules for a box in an orthogonal flow [CSS3-WRITING-MODES]. The flex base size is the item’s max-content main size.
    5. Otherwise, lay out the item into the available space using its used flex basis in place of its main size, treating a main size of auto as max-content. If a cross size is needed to determine the main size (e.g. when the flex item’s main size is in its block axis) and the flex item’s cross size is auto and not definite, in this calculation use fit-content as the flex item’s cross size. The flex base size is the item’s resulting main size.

    The hypothetical main size is the item’s flex base size clamped according to its min and max main size properties.

  3. Determine the main size of the flex container using the rules of the formatting context in which it participates. For this computation, auto margins on flex items are treated as 0.

9.3 Main Size Determination

  1. Collect flex items into flex lines:
  2. Resolve the flexible lengths of all the flex items to find their used main size (see section 9.7.).

9.4 Cross Size Determination

  1. Determine the hypothetical cross size of each item by performing layout with the used main size and the available space, treating auto as fit-content.
  2. Calculate the cross size of each flex line.

    If the flex container has only one flex line (even if it’s a multi-line flex container) and has a definite cross size, the cross size of the flex line is the flex container’s inner cross size.

    Otherwise, for each flex line:

    1. Collect all the flex items whose inline-axis is parallel to the main-axis, whose align-self is baseline, and whose cross-axis margins are both non-auto. Find the largest of the distances between each item’s baseline and its hypothetical outer cross-start edge, and the largest of the distances between each item’s baseline and its hypothetical outer cross-end edge, and sum these two values.
    2. Among all the items not collected by the previous step, find the largest outer hypothetical cross size.
    3. The used cross-size of the flex line is the largest of the numbers found in the previous two steps and zero.

      If the flex container has only one flex line (even if it’s a multi-line flex container), then clamp the line’s cross-size to be within the container’s computed min and max cross-size properties. Note that if CSS 2.1’s definition of min/max-width/height applied more generally, this behavior would fall out automatically.

  3. Handle 'align-content: stretch'. If the flex container has a definite cross size, align-content is stretch, and the sum of the flex lines' cross sizes is less than the flex container’s inner cross size, increase the cross size of each flex line by equal amounts such that the sum of their cross sizes exactly equals the flex container’s inner cross size.
  4. Collapse visibility:collapse items. If any flex items have visibility: collapse, note the cross size of the line they’re in as the item’s strut size, and restart layout from the beginning.

    In this second layout round, when collecting items into lines, treat the collapsed items as having zero main size. For the rest of the algorithm following that step, ignore the collapsed items entirely (as if they were display:none) except that after calculating the cross size of the lines, if any line’s cross size is less than the largest strut size among all the collapsed items in the line, set its cross size to that strut size.

    Skip this step in the second layout round.

  5. Determine the used cross size of each flex item. If a flex item has align-self: stretch, its computed cross size property is auto, and neither of its cross-axis margins are auto, the used outer cross size is the used cross size of its flex line, clamped according to the item’s min and max cross size properties. Otherwise, the used cross size is the item’s hypothetical cross size.

    If the flex item has align-self: stretch, redo layout for its contents, treating this used size as its definite cross size so that percentage-sized children can be resolved.

    Note that this step does not affect the main size of the flex item, even if it has an intrinsic aspect ratio.

9.5 Main-Axis Alignment

  1. Distribute any remaining free space. For each flex line:
    1. If the remaining free space is positive and at least one main-axis margin on this line is auto, distribute the free space equally among these margins. Otherwise, set all auto margins to zero.
    2. Align the items along the main-axis per justify-content.

9.6 Cross-Axis Alignment

  1. Resolve cross-axis auto margins. If a flex item has auto cross-axis margins:
  2. Align all flex items along the cross-axis per align-self, if neither of the item’s cross-axis margins are auto.
  3. Determine the flex container’s used cross size:
  4. Align all flex lines per align-content.

9.7 Resolving Flexible Lengths

To resolve the flexible lengths of the items within a flex line:

  1. Determine the used flex factor. Sum the outer hypothetical main sizes of all items on the line. If the sum is less than the flex container’s inner main size, use the flex grow factor for the rest of this algorithm; otherwise, use the flex shrink factor.
  2. Size inflexible items. Freeze, setting its target main size to its hypothetical main size
  3. Calculate initial free space. Sum the outer sizes of all items on the line, and subtract this from the flex container’s inner main size. For frozen items, use their outer target main size; for other items, use their outer flex base size.
  4. Loop:
    1. Check for flexible items. If all the flex items on the line are frozen, free space has been distributed; exit this loop.
    2. Calculate the remaining free space as for initial free space, above. If the sum of the unfrozen flex items’ flex factors is less than one, multiply the initial free space by this sum. If the magnitude of this value is less than the magnitude of the remaining free space, use this as the remaining free space.
    3. Distribute free space proportional to the flex factors.
      If the remaining free space is zero
      Do nothing.
      If using the flex grow factor
      Find the ratio of the item’s flex grow factor to the sum of the flex grow factors of all unfrozen items on the line. Set the item’s target main size to its flex base size plus a fraction of the remaining free space proportional to the ratio.
      If using the flex shrink factor
      For every unfrozen item on the line, multiply its flex shrink factor by its outer flex base size, and note this as its scaled flex shrink factor. Find the ratio of the item’s scaled flex shrink factor to the sum of the scaled flex shrink factors of all unfrozen items on the line. Set the item’s target main size to its flex base size minus a fraction of the absolute value of the remaining free space proportional to the ratio. Note this may result in a negative inner main size; it will be corrected in the next step.
      Otherwise
      Do nothing.
    4. Fix min/max violations. Clamp each non-frozen item’s target main size by its min and max main size properties. If the item’s target main size was made smaller by this, it’s a max violation. If the item’s target main size was made larger by this, it’s a min violation.
    5. Freeze over-flexed items. The total violation is the sum of the adjustments from the previous step ∑(clamped size - unclamped size). If the total violation is:
      Zero
      Freeze all items.
      Positive
      Freeze all the items with min violations.
      Negative
      Freeze all the items with max violations.
    6. Return to the start of this loop.
  5. Set each item’s used main size to its target main size.

9.8 Definite and Indefinite Sizes

For the purposes of these definitions, a definite size is one that can be determined without measuring content, i.e. is a <length>, a size of the initial containing block, or a <percentage> that is resolved against a definite size. An indefinite size is one that is not definite.

Flexbox has several additional cases where a length can be considered definite:

  1. If a single-line flex container has a definite cross size, the outer cross size of any stretched flex items is the flex container’s inner cross size (clamped to the flex item’s min and max cross size) and is considered definite.
  2. If a percentage is going to be resolved against a flex item’s main size, and the flex item has a definite flex basis, the main size must be treated as definite for the purpose of resolving the percentage, and the percentage must resolve against the flexed main size of the flex item (that is, after the layout algorithm below has been completed for the flex item’s flex container, and the flex item has acquired its final size).

9.9 Intrinsic Sizes

The max-content main size of a flex container is the smallest size the flex container can take while maintaining the max-content contributions of its flex items.

  1. For each flex item, subtract its flex basis from its max-content contribution size, then divide by its flex grow factor, floored at 1, or by its scaled flex shrink factor (if the result was negative, flooring the flex shrink factor at 1 if necessary). This is the item’s max-content flex fraction.
  2. Place all flex items into lines of infinite length.
  3. Within each line, find the largest max-content flex fraction among all the flex items. Add each item’s flex basis to the product of its flex grow factor (or scaled flex shrink factor, if the chosen max-content flex fraction was negative) and the chosen max-content flex fraction, then clamp that result by the max main size property.
  4. The flex container’s max-content size is the largest sum of the afore-calculated sizes of all items within a single line.

The min-content main size of a single-line flex container is calculated identically to the max-content main size, except that the flex item’s min-content contribution is used instead of its max-content contribution. For a multi-line container, it is simply the largest min-content contribution of all the flex items in the flex container.

The min-content cross size and max-content cross size of a flex container are the cross size of the flex container after performing layout into the given available main-axis space and infinite available cross-axis space.

The main-size min-content/max-content contribution of a flex item is its outer hypothetical main size when sized under a min-content/max-content constraint (respectively).

See [CSS3-SIZING] for a definition of the terms in this section.

10 Fragmenting Flex Layout

Flex containers can break across pages between items, between lines of items (in multi-line mode), and inside items. The break-* properties apply to flex containers as normal for block-level or inline-level boxes. This section defines how they apply to flex items and the contents of flex items.

The following breaking rules refer to the fragmentation container as the “page”. The same rules apply to any other fragmenters. (Substitute “page” with the appropriate fragmenter type as needed.) See the CSS3 Fragmentation Module [CSS3-BREAK]. For readability, in this section the terms "row" and "column" refer to the relative orientation of the flex container with respect to the block flow direction of the fragmentation context, rather than to that of the flex container itself.

The exact layout of a fragmented flex container is not defined in this level of Flexible Box Layout. However, breaks inside a flex container are subject to the following rules:

10.1 Sample Flex Fragmentation Algorithm

This informative section presents a possible fragmentation algorithm for flex containers. Implementors are encouraged to improve on this algorithm and provide feedback to the CSS Working Group.

This algorithm assumes that pagination always proceeds only in the forward direction; therefore, in the algorithms below, alignment is mostly ignored prior to pagination. Advanced layout engines may be able to honor alignment across fragments.

Single-line column flex container
  1. Run the flex layout algorithm (without regards to pagination) through Cross Sizing Determination.
  2. Lay out as many consecutive flex items or item fragments as possible (but at least one or a fragment thereof), starting from the first, until there is no more room on the page or a forced break is encountered.
  3. If the previous step ran out of room and the free space is positive, the UA may reduce the distributed free space on this page (down to, but not past, zero) in order to make room for the next unbreakable flex item or fragment. Otherwise, the item or fragment that does not fit is pushed to the next page. The UA should pull up if more than 50% of the fragment would have fit in the remaining space and should push otherwise.
  4. If there are any flex items or fragments not laid out by the previous steps, rerun the flex layout algorithm from Line Length Determination through Cross Sizing Determination with the next page’s size and all the contents (including those already laid out), and return to the previous step, but starting from the first item or fragment not already laid out.
  5. For each fragment of the flex container, continue the flex layout algorithm from Main-Axis Alignment to its finish.

It is the intent of this algorithm that column-direction single-line flex containers paginate very similarly to block flow. As a test of the intent, a flex container with justify-content:start and no flexible items should paginate identically to a block with in-flow children with same content, same used size and same used margins.

Multi-line column flex container
  1. Run the flex layout algorithm with regards to pagination (limiting the flex container’s maximum line length to the space left on the page) through Cross Sizing Determination.
  2. Lay out as many flex lines as possible (but at least one) until there is no more room in the flex container in the cross dimension or a forced break is encountered:
    1. Lay out as many consecutive flex items as possible (but at least one), starting from the first, until there is no more room on the page or a forced break is encountered. Forced breaks within flex items are ignored.
    2. If this is the first flex container fragment, this line contains only a single flex item that is larger than the space left on the page, and the flex container is not at the top of the page already, move the flex container to the next page and restart flex container layout entirely.
    3. If there are any flex items not laid out by the first step, rerun the flex layout algorithm from Main Sizing Determination through Cross Sizing Determination using only the items not laid out on a previous line, and return to the previous step, starting from the first item not already laid out.
  3. If there are any flex items not laid out by the previous step, rerun the flex layout algorithm from Line Sizing Determination through Cross Sizing Determination with the next page’s size and only the items not already laid out, and return to the previous step, but starting from the first item not already laid out.
  4. For each fragment of the flex container, continue the flex layout algorithm from Main-Axis Alignment to its finish.

If a flex item does not entirely fit on a single page, it will not be paginated in multi-line column flex containers.

Single-line row flex container
  1. Run the entire flex layout algorithm (without regards to pagination), except treat any align-self other than start or baseline as start.
  2. If an unbreakable item doesn’t fit within the space left on the page, and the flex container is not at the top of the page, move the flex container to the next page and restart flex container layout entirely.
  3. For each item, lay out as much of its contents as will fit in the space left on the page, and fragment the remaining content onto the next page, rerunning the flex layout algorithm from Line Length Determination through Main-Axis Alignment into the new page size using all the contents (including items completed on previous pages).

    Any flex items that fit entirely into previous fragments still take up space in the main axis in later fragments.

  4. For each fragment of the flex container, rerun the flex layout algorithm from Cross-Axis Alignment to its finish. For all fragments besides the first, treat align-self and align-content as being start for all item fragments and lines.
  5. If any item, when aligned according to its original align-self value into the combined cross size of all the flex container fragments, would fit entirely within a single flex container fragment, it may be shifted into that fragment and aligned appropriately.
Multi-line row flex container
  1. Run the flex layout algorithm (without regards to pagination), through Cross Sizing Determination.
  2. Lay out as many flex lines as possible (but at least one), starting from the first, until there is no more room on the page or a forced break is encountered.

    If a line doesn’t fit on the page, and the line is not at the top of the page, move the line to the next page and restart the flex layout algorithm entirely, using only the items in and following this line.

    If a flex item itself causes a forced break, rerun the flex layout algorithm from Main Sizing Determination through Main-Axis Alignment, using only the items on this and following lines, but with the item causing the break automatically starting a new line in the line breaking step, then continue with this step. Forced breaks within flex items are ignored.

  3. If there are any flex items not laid out by the previous step, rerun the flex layout algorithm from Line Length Determination through Main-Axis Alignment with the next page’s size and only the items not already laid out. Return to the previous step, but starting from the first line not already laid out.
  4. For each fragment of the flex container, continue the flex layout algorithm from Cross Axis Alignment to its finish.

Acknowledgments

Thanks for feedback and contributions to

Erik Anderson, Tony Chang, Phil Cupp, Arron Eicholz, James Elmore, Andrew Fedoniouk, Brian Heuston, Shinichiro Hamaji, Daniel Holbert, Ben Horst, John Jansen, Brad Kemper, Kang-hao Lu, Markus Mielke, Peter Moulder, Robert O’Callahan, Christoph Päper, Ning Rogers, Peter Salas, Morten Stenshorne, Christian Stockwell, Ojan Vafai, Eugene Veselov, Greg Whitworth, Boris Zbarsky.

Changes

Changes since the 25 March 2014 LCWD

This section documents the changes since the previous publication. A Disposition of Comments is also available.

Substantive Changes and Bugfixes

The following significant changes were made since the 25 March 2014 Last Call Working Draft

Clarifications

The following significant changes were made since the 25 March 2014 Last Call Working Draft

Changes since the 18 September 2012 Candidate Recommendation

This section documents the changes since the previous publication. A Disposition of Comments is also available.

Substantive Changes and Bugfixes

The following significant changes were made since the 18 September 2012 Candidate Recommendation:

Clarifications

The following significant clarifications were also made:

A Disposition of Candidate Recommendation Comments is available.

Conformance

Document conventions

Conformance requirements are expressed with a combination of descriptive assertions and RFC 2119 terminology. The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in the normative parts of this document are to be interpreted as described in RFC 2119. However, for readability, these words do not appear in all uppercase letters in this specification.

All of the text of this specification is normative except sections explicitly marked as non-normative, examples, and notes. [RFC2119]

Examples in this specification are introduced with the words "for example" or are set apart from the normative text with class="example", like this:

This is an example of an informative example.

Informative notes begin with the word "Note" and are set apart from the normative text with class="note", like this:

Note, this is an informative note.

Advisements are normative sections styled to evoke special attention and are set apart from other normative text with <strong class="advisement">, like this: UAs MUST provide an accessible alternative.

Conformance classes

Conformance to this specification is defined for three conformance classes:

style sheet
A CSS style sheet.
renderer
A UA that interprets the semantics of a style sheet and renders documents that use them.
authoring tool
A UA that writes a style sheet.

A style sheet is conformant to this specification if all of its statements that use syntax defined in this module are valid according to the generic CSS grammar and the individual grammars of each feature defined in this module.

A renderer is conformant to this specification if, in addition to interpreting the style sheet as defined by the appropriate specifications, it supports all the features defined by this specification by parsing them correctly and rendering the document accordingly. However, the inability of a UA to correctly render a document due to limitations of the device does not make the UA non-conformant. (For example, a UA is not required to render color on a monochrome monitor.)

An authoring tool is conformant to this specification if it writes style sheets that are syntactically correct according to the generic CSS grammar and the individual grammars of each feature in this module, and meet all other conformance requirements of style sheets as described in this module.

Partial implementations

So that authors can exploit the forward-compatible parsing rules to assign fallback values, CSS renderers must treat as invalid (and ignore as appropriate) any at-rules, properties, property values, keywords, and other syntactic constructs for which they have no usable level of support. In particular, user agents must not selectively ignore unsupported component values and honor supported values in a single multi-value property declaration: if any value is considered invalid (as unsupported values must be), CSS requires that the entire declaration be ignored.

Experimental implementations

To avoid clashes with future CSS features, the CSS2.1 specification reserves a prefixed syntax for proprietary and experimental extensions to CSS.

Prior to a specification reaching the Candidate Recommendation stage in the W3C process, all implementations of a CSS feature are considered experimental. The CSS Working Group recommends that implementations use a vendor-prefixed syntax for such features, including those in W3C Working Drafts. This avoids incompatibilities with future changes in the draft.

Non-experimental implementations

Once a specification reaches the Candidate Recommendation stage, non-experimental implementations are possible, and implementors should release an unprefixed implementation of any CR-level feature they can demonstrate to be correctly implemented according to spec.

To establish and maintain the interoperability of CSS across implementations, the CSS Working Group requests that non-experimental CSS renderers submit an implementation report (and, if necessary, the testcases used for that implementation report) to the W3C before releasing an unprefixed implementation of any CSS features. Testcases submitted to W3C are subject to review and correction by the CSS Working Group.

Further information on submitting testcases and implementation reports can be found from on the CSS Working Group’s website at http://www.w3.org/Style/CSS/Test/. Questions should be directed to the public-css-testsuite@w3.org mailing list.

References

Normative References

[CSS21]
Bert Bos; et al. Cascading Style Sheets Level 2 Revision 1 (CSS 2.1) Specification. 7 June 2011. W3C Recommendation. URL: http://www.w3.org/TR/2011/REC-CSS2-20110607
[CSS3-BREAK]
Rossen Atanassov; Elika J. Etemad. CSS Fragmentation Module Level 3. 23 August 2012. W3C Working Draft. (Work in progress.) URL: http://www.w3.org/TR/2012/WD-css3-break-20120823/
[CSS3-SIZING]
Tab Atkins Jr.; Elika J. Etemad. CSS Intrinsic & Extrinsic Sizing Module Level 3. 27 September 2012. W3C Working Draft. (Work in progress.) URL: http://www.w3.org/TR/2012/WD-css3-sizing-20120927/
[CSS3-WRITING-MODES]
Elika J. Etemad; Koji Ishii. CSS Writing Modes Module Level 3. 15 November 2012. W3C Working Draft. (Work in progress.) URL: http://www.w3.org/TR/2012/WD-css3-writing-modes-20121115/
[RFC2119]
S. Bradner. Key words for use in RFCs to Indicate Requirement Levels. URL: http://www.ietf.org/rfc/rfc2119.txt

Informative References

[CSS3-ALIGN]
Elika J. Etemad; Tab Atkins Jr.. CSS Box Alignment Module Level 3. 14 May 2013. W3C Working Draft. (Work in progress.) URL: http://www.w3.org/TR/2013/WD-css3-align-20130514/
[CSS3UI]
Tantek Çelik. CSS Basic User Interface Module Level 3 (CSS3 UI). 17 January 2012. W3C Working Draft. (Work in progress.) URL: http://www.w3.org/TR/2012/WD-css3-ui-20120117/
[CSS3VAL]
Håkon Wium Lie; Tab Atkins; Elika J. Etemad. CSS Values and Units Module Level 3. 30 July 2013. W3C Candidate Recommendation. (Work in progress.) URL: http://www.w3.org/TR/2013/CR-css3-values-20130730/
[HTML40]
Ian Jacobs; David Raggett; Arnaud Le Hors. HTML 4.0 Specification. 24 April 1998. W3C Recommendation. URL: http://www.w3.org/TR/1998/REC-html40-19980424

Index

Property index

NameValueInitialApplies toInh.%agesMediaAnimatableCanonical orderComputed value
flex-directionrow | row-reverse | column | column-reverserowflex containersnovisualspecified value
flex-wrapnowrap | wrap | wrap-reversenowrapflex containersnovisualspecified value
flex-flow<flex-direction> || <flex-wrap>see individual propertiesflex containerssee individual propertiesvisualsee individual properties
order<integer>0flex items and absolutely-positioned children of flex containersnovisualyesspecified value
flexnone | auto | [ <‘flex-grow’> <‘flex-shrink’>? || <‘flex-basis’> ]see individual propertiesflex itemssee individual propertiesvisualper grammarsee individual properties
flex-grow<number>0flex itemsnovisualyesspecified value
flex-shrink<number>1flex itemsnovisualyesspecified value
flex-basismain-size | <‘width’>main-sizeflex itemsnorelative to the flex container’s inner main sizevisualas widthas specified, with lengths made absolute
justify-contentflex-start | flex-end | center | space-between | space-aroundflex-startflex containersnovisualspecified value
align-itemsflex-start | flex-end | center | baseline | stretchstretchflex containersnovisualspecified value
align-selfauto | flex-start | flex-end | center | baseline | stretchautoflex itemsnovisualauto computes to parent’s align-items value; otherwise as specified
align-contentflex-start | flex-end | center | space-between | space-around | stretchstretchmulti-line flex containersnovisualspecified value

Issues Index

This value was originally named auto, but that was problematic because it left us with no keyword to represent an automatic flex basis. There were two options for fixing this:
auto for automatic sizing; main-axis for size lookup
The advantage of this is primarily that auto means “automatic” both when it is specified for flex-basis and when it is pulled in from width/height. It has the disadvantage (due to back-compat concerns) that the flex shorthand had to be tweaked to expand flex: auto to flex: auto main-size rather than to flex: 1 auto. It is the currently-implemented option, but may be unshippable due to back-compat constraints.
auto for size lookup; content1 for automatic sizing
The advantage of this is that it will not break existing content: a flex-basis keyword for automatic sizing did not exist in the original CR, so this option adds a new keyword but does not rename existing functionality. It therefore avoids the flex: autoflex: 1 auto inequality.

The CSSWG is waiting for feedback from implementers on this issue.

1 Alternate keyword suggestions welcome.