The Anchor Class

/*
**	(c) COPYRIGHT MIT 1995.
**	Please first read the full copyright statement in the file COPYRIGH.
*/

An anchor represents a region of a hypertext document which is linked to another anchor in the same or a different document. Another name for anchors would be URLs as an anchor represents all we know about a URL - including where it points to and who points to it. Because the anchor objects represent the part of the Web, the application has been in touch, it is often useful to maintain the anchors throughout the lifetime of the application. It would actually be most useful if we had persistent anchors so that an application could build up a higher knowledge about the Web topology.

When to Escape and Unescape Addresses

The URI escape policy in libwww is that all URIs created as anchors must already have been escaped. The reason for this is that if URIs are not escaped then the URI parser is not guaranteed to work as expected. Imagine, for example, that you have a ":" in a host name, then you could get something like this: http://my:host:8000/ instead of http://my%3Ahost:8000/.

Libwww provides support for escaping and unescaping URIs using this set of APIs.

This module is implemented by HTAnchor.c, and it is a part of the W3C Sample Code Library.

#ifndef HTANCHOR_H
#define HTANCHOR_H

Types defined and used by the Anchor Object

This is a set of videly used type definitions used through out the Library:

#include "WWWUtil.h"

typedef HTAtom * HTFormat;
typedef HTAtom * HTLevel;		       /* Used to specify HTML level */
typedef HTAtom * HTEncoding;				    /* C-E and C-T-E */
typedef HTAtom * HTCharset;
typedef HTAtom * HTLanguage;

typedef struct _HTAnchor	HTAnchor;
typedef struct _HTParentAnchor	HTParentAnchor;
typedef struct _HTChildAnchor	HTChildAnchor;

#include "HTLink.h"
#include "HTMethod.h"
#include "HTResponse.h"

The Anchor Class

We have three variants of the Anchor object - I guess some would call them superclass and subclasses ;-)

Anchor Base Class

This is the super class of anchors. We often use this as an argument to the functions that both accept parent anchors and child anchors. We separate the first link from the others to avoid too many small mallocs involved by a list creation. Most anchors only point to one place.

Anchor for a Parent Object

These anchors points to the whole contents of any resource accesible by a URI. The parent anchor now contains all known metainformation about that object and in some cases the parent anchor also contains the document itself. Often we get the metainformation about a document via the entity headers in the HTTP specification.

Anchor for a Child Object

A child anchor is a anchor object that points to a subpart of a hypertext document. In HTML this is represented by the NAME tag of the Anchor element.

After we have defined the data structures we must define the methods that can be used on them. All anchors are kept in an internal hash table so that they are easier to find again.

Find/Create a Parent Anchor

This one is for a reference (link) which is found in a document, and might not be already loaded. The parent anchor returned can either be created on the spot or is already in the hash table.

extern HTAnchor * HTAnchor_findAddress		(const char * address);

Find/Create a Child Anchor

This one is for a new child anchor being edited into an existing document. The parent anchor must already exist but the child returned can either be created on the spot or is already in the hash table. The tag is the part that's after the '#' sign in a URI.

extern HTChildAnchor * HTAnchor_findChild	(HTParentAnchor *parent,
						 const char *	tag);

Find/Create a Child Anchor and Link to Another Parent

Find a child anchor anchor with a given parent and possibly a tag, and (if passed) link this child to the URI given in the href. As we really want typed links to the caller should also indicate what the type of the link is (see HTTP spec for more information). The link is relative to the address of the parent anchor.

extern HTChildAnchor * HTAnchor_findChildAndLink (
		HTParentAnchor * parent,		/* May not be 0 */
		const char * tag,			/* May be "" or 0 */
		const char * href,			/* May be "" or 0 */
		HTLinkType ltype);			/* May be 0 */

Delete an Anchor

All outgoing links from parent and children are deleted, and this anchor is removed from the sources list of all its targets. We also delete the targets. If this anchor's source list is empty, we delete it and its children.

extern BOOL HTAnchor_delete	(HTParentAnchor *me);

Clear all Anchors

Deletes all the metadata associated with anchors but doesn't delete the anchor link structure itself. This is much safer than deleting the complete anchor structure as this represents the complete Web the application has been in touch with. It also returns a list of all the objects (hyperdoc) hanging of the parent anchors found while doing it. These are not deleted by libwww.

extern BOOL HTAnchor_clearAll (HTList * documents);

Delete all Anchors

Deletes all anchors and return a list of all the objects (hyperdoc) hanging of the parent anchors found while doing it. The application may keep its own list of HyperDocs, but this function returns it anyway. It is always for the application to delete any HyperDocs. If NULL then no hyperdocs are returned. Return YES if OK, else NO.

Note: This function is different from cleaning up the history list!

extern BOOL HTAnchor_deleteAll	(HTList * objects);

Flatten all anchors into Array

Flattens the anchor web structure into an array. This is useful for calculating statistics, sorting the parent anchors etc.

The caller can indicate the size of the array (total number of anchors if known - otherwise 0).

Return an array that must be freed by the caller or NULL if no anchors.

extern HTArray * HTAnchor_getArray (int growby);

Links and Anchors

Anchor objects are bound together by Link objects that carry information about what type of link and whetther we have followed the link etc. Any anchor object can have zero, one, or many links but the normal case is one. Therefore we treat this is a special way.

Handling the Main Link

Any outgoing link can at any time be the main destination.

extern BOOL HTAnchor_setMainLink	(HTAnchor * anchor, HTLink * link);
extern HTLink * HTAnchor_mainLink	(HTAnchor * anchor);

extern HTAnchor * HTAnchor_followMainLink (HTAnchor * anchor);

Handling the Sub Links

extern BOOL HTAnchor_setSubLinks	(HTAnchor * anchor, HTList * list);
extern HTList * HTAnchor_subLinks	(HTAnchor * anchor);

Search for a Link Type

Links can have relations (indicated by the "rel" or "rev" HTML link attributes). This function can search an anchor looking for a specific type, for example "stylesheet".
extern HTLink * HTAnchor_findLinkType (HTAnchor * me, HTLinkType type);

Relations Between Children and Parents

As always, children and parents have a compliated relationship and the libwww Anchor class is no exception.

Who is Parent?

For parent anchors this returns the anchor itself

extern HTParentAnchor * HTAnchor_parent	(HTAnchor *me);

Does it have any Anchors within it?

Does this parent anchor have any children

extern BOOL HTAnchor_hasChildren	(HTParentAnchor *me);

Is this anchor a Child?

extern BOOL HTAnchor_isChild (HTAnchor * me);

Get the Tag/Fragment/View of this anchor

If this is a child anchor then it has a tag (often also called a "fragment"), which is essentially a specific view of a document. This is why I like to call it a view instead of a fragment. The string returned (if non-NULL) must be freed by the caller.
extern char * HTAnchor_view (HTAnchor * me);

Anchor Addresses

There are two addresses of an anchor. The URI that was passed when the anchor was crated and the physical address that's used when the URI is going to be requested. The two addresses may be different if the request is going through a proxy or a gateway or it may have been mapped through a rule file.

Logical Address

Returns the full URI of the anchor, child or parent as a malloc'd string to be freed by the caller as when the anchor was created.

extern char * HTAnchor_address		(HTAnchor * me);

Expanded Logical Address

When expanding URLs within a hypertext document, the base address is taken as the following value if present (in that order):

extern char * HTAnchor_expandedAddress  (HTAnchor * me);

Physical address

Contains the physical address after we haved looked for proxies etc.

extern char * HTAnchor_physical		(HTParentAnchor * me);
extern void HTAnchor_setPhysical	(HTParentAnchor * me, char * protocol);
extern void HTAnchor_clearPhysical	(HTParentAnchor * me);

Entity Body Information

A parent anchor can have a data object bound to it. This data object does can for example be a parsed version of a HTML that knows how to present itself to the user, or it can be an unparsed data object. It's completely free for the application to use this possibility, but a typical usage would to manage the data object as part of a memory cache.

extern void HTAnchor_setDocument	(HTParentAnchor *me, void * doc);
extern void * HTAnchor_document		(HTParentAnchor *me);

Entity Header Information

The anchor object also contains all the metainformation that we know about the object.

Clear All header Information

extern void HTAnchor_clearHeader	(HTParentAnchor *me);

Inherit Metainformation from the Response object

Once we have decided to cache the object we transfer already parsed metainformation from the HTResponse object to the anchor object and also the unparsed headers as we may wanna use that information later.

extern BOOL HTAnchor_update (HTParentAnchor * me, HTResponse * response);

Is the Anchor searchable?

extern void HTAnchor_clearIndex		(HTParentAnchor * me);
extern void HTAnchor_setIndex		(HTParentAnchor * me);
extern BOOL HTAnchor_isIndex		(HTParentAnchor * me);

Anchor Title

We keep the title in the anchor as we then can refer to it later in the history list etc. We can also obtain the title element if it is passed as a HTTP header in the response. Any title element found in an HTML document will overwrite a title given in a HTTP header.

extern const char * HTAnchor_title	(HTParentAnchor *me);
extern void HTAnchor_setTitle		(HTParentAnchor *me,
					 const char *	title);
extern void HTAnchor_appendTitle	(HTParentAnchor *me,
					 const char *	title);

Meta Tags within the Document

extern HTAssocList * HTAnchor_meta (HTParentAnchor * me);
extern BOOL HTAnchor_addMeta (HTParentAnchor * me,
			      const char * name, const char * value);

The Robots Meta tag

A special case function that looks for any robots meta tag. This tag contains information about which links a robot can traverse and which it shouldn't.
extern char * HTAnchor_robots (HTParentAnchor * me);

Content Base

The Content-Base header may be used for resolving relative URLs within the entity. If it there is no Content-Base header then we use the Content-Location if present and absolute.

extern char * HTAnchor_base	(HTParentAnchor * me);
extern BOOL HTAnchor_setBase 	(HTParentAnchor * me, char * base);

Content Location

Content location can either be an absolute or a relative URL. The URL may be either absolute or relative. If it is relative then we parse it relative to the Content-Base header of the request URI if any, otherwise we use the Request URI.

extern char * HTAnchor_location		(HTParentAnchor * me);
extern BOOL HTAnchor_setLocation	(HTParentAnchor * me, char * location);

Media Types (Content-Type)

extern HTFormat HTAnchor_format		(HTParentAnchor *me);
extern void HTAnchor_setFormat		(HTParentAnchor *me,
					 HTFormat	form);

Content Type Parameters

The Anchor obejct stores all content parameters in an Association list so here you will always be able to find them. We also have a few methods for the special cases: charset and level as they are often needed.

extern HTAssocList * HTAnchor_formatParam (HTParentAnchor * me);

extern BOOL HTAnchor_addFormatParam 	(HTParentAnchor * me,
					const char * name, const char * value);

Charset parameter to Content-Type

extern HTCharset HTAnchor_charset	(HTParentAnchor *me);
extern BOOL HTAnchor_setCharset		(HTParentAnchor *me,
					 HTCharset	charset);

Level parameter to Content-Type

extern HTLevel HTAnchor_level		(HTParentAnchor * me);
extern BOOL HTAnchor_setLevel		(HTParentAnchor * me,
					 HTLevel	level);

Content Language

extern HTList * HTAnchor_language	(HTParentAnchor * me);
extern BOOL HTAnchor_addLanguage	(HTParentAnchor *me, HTLanguage lang);
extern BOOL HTAnchor_deleteLanguageAll  (HTParentAnchor * me);

Content Encoding

extern HTList * HTAnchor_encoding	(HTParentAnchor * me);
extern BOOL HTAnchor_addEncoding	(HTParentAnchor * me, HTEncoding enc);
extern BOOL HTAnchor_deleteEncoding     (HTParentAnchor * me, HTEncoding enc);
extern BOOL HTAnchor_deleteEncodingAll  (HTParentAnchor * me);

#define HTAnchor_removeEncoding(a, e)   HTAnchor_deleteEncoding((a), (e))

Content Transfer Encoding

extern HTEncoding HTAnchor_contentTransferEncoding	(HTParentAnchor *me);
extern void HTAnchor_setContentTransferEncoding	        (HTParentAnchor *me,
				 	                 HTEncoding	cte);

Content Length

extern long int HTAnchor_length	(HTParentAnchor * me);
extern void HTAnchor_setLength	(HTParentAnchor * me, long int length);
extern void HTAnchor_addLength	(HTParentAnchor * me, long int deltalength);

Content MD5

extern char * HTAnchor_md5	(HTParentAnchor * me);
extern BOOL HTAnchor_setMd5	(HTParentAnchor * me, const char * hash);

Allowed methods (Allow)

extern HTMethod HTAnchor_allow   (HTParentAnchor * me);
extern void HTAnchor_setAllow    (HTParentAnchor * me, HTMethod methodset);
extern void HTAnchor_appendAllow (HTParentAnchor * me, HTMethod methodset);

Version

extern char * HTAnchor_version	(HTParentAnchor * me);
extern void HTAnchor_setVersion	(HTParentAnchor * me, const char * version);

Date

Returns the date that was registered in the RFC822 header "Date"

extern time_t HTAnchor_date		(HTParentAnchor * me);
extern void HTAnchor_setDate		(HTParentAnchor	* me, const time_t date);

Last Modified Date

Returns the date that was registered in the RFC822 header "Last-Modified"

extern time_t HTAnchor_lastModified	(HTParentAnchor * me);
extern void HTAnchor_setLastModified	(HTParentAnchor	* me, const time_t lm);

Entity Tag

Entity tags are used for comparing two or more entities from the same requested resource. It is a cache validator much in the same way Date can be. The difference is that we can't always trust the date and time stamp and hence we must have something stronger.

extern char * HTAnchor_etag 	(HTParentAnchor * me);
extern void HTAnchor_setEtag	(HTParentAnchor * me, const char * etag);
extern BOOL HTAnchor_isEtagWeak	(HTParentAnchor * me);

Age Header

The Age response-header field conveys the sender's estimate of the amount of time since the response (or its revalidation) was generated at the origin server. A cached response is "fresh" if its age does not exceed its freshness lifetime.

extern time_t HTAnchor_age    (HTParentAnchor * me);
extern void HTAnchor_setAge   (HTParentAnchor * me, const time_t age);

Expires Date

extern time_t HTAnchor_expires		(HTParentAnchor * me);
extern void HTAnchor_setExpires		(HTParentAnchor	* me, const time_t exp);

Derived from

extern char * HTAnchor_derived	(HTParentAnchor *me);
extern void HTAnchor_setDerived	(HTParentAnchor *me, const char *derived_from);

Status of Header Parsing

This is primarily for internal use. It is so that we can check whether the header has been parsed or not.

extern BOOL HTAnchor_headerParsed	(HTParentAnchor *me);
extern void HTAnchor_setHeaderParsed	(HTParentAnchor *me);

Original Response Headers

The MIME parser may add the original response headers as (name,value) pairs.

extern BOOL HTAnchor_setHeader       (HTParentAnchor * me, HTAssocList * list);
extern HTAssocList * HTAnchor_header (HTParentAnchor * me);
#endif /* HTANCHOR_H */


@(#) $Id: HTAnchor.html,v 2.61 2000/08/04 09:14:58 kahan Exp $