Annotation of libwww/Library/src/HTFormat.html, revision 2.7

2.1       timbl       1: <HEADER>
                      2: <TITLE>HTFormat: The format manager in the WWW Library</TITLE>
2.7     ! timbl       3: <NEXTID N="10">
2.1       timbl       4: </HEADER>
                      5: <BODY>
                      6: <H1>Manage different document formats</H1>Here we describe the functions of
                      7: the HTFormat module which handles
                      8: conversion between different data
                      9: representations.  (In MIME parlance,
                     10: a representation is known as a content-type.
2.2       timbl      11: In WWW  the term "format" is often
2.1       timbl      12: used as it is shorter).<P>
                     13: This module is implemented by <A
2.7     ! timbl      14: NAME=z0 HREF="HTFormat.c">HTFormat.c</A>
        !            15: . This hypertext document is used
        !            16: to generate the <A
        !            17: NAME=z8 HREF="HTFormat.h">HTFormat.h</A> inlude
        !            18: file.
2.1       timbl      19: <H2>Preamble</H2>
                     20: <PRE>#ifndef HTFORMAT_H
                     21: #define HTFORMAT_H
                     22: 
                     23: #include "HTUtils.h"
                     24: #include <A
                     25: NAME=z7 HREF="HTStream.html">"HTStream.h"</A>
                     26: #include "HTAtom.h"
2.2       timbl      27: #include "HTList.h"
2.1       timbl      28: 
                     29: #ifdef SHORT_NAMES
                     30: #define HTOutputSource HTOuSour
                     31: #define HTOutputBinary HTOuBina
                     32: #endif
                     33: 
                     34: </PRE>
                     35: <H2>The HTFormat type</H2>We use the HTAtom object for holding
                     36: representations. This allows faster
                     37: manipulation (comparison and copying)
                     38: that if we stayed with strings.
                     39: <PRE>typedef HTAtom * HTFormat;
                     40:                        
                     41: </PRE>These macros (which used to be constants)
                     42: define some basic internally referenced
2.2       timbl      43: representations.  The www/xxx ones
2.1       timbl      44: are of course not MIME standard.<P>
                     45: www/source  is an output format which
                     46: leaves the input untouched. It is
                     47: useful for diagnostics, and for users
                     48: who want to see the original, whatever
                     49: it is.
                     50: <PRE>                  /* Internal ones */
                     51: #define WWW_SOURCE HTAtom_for("www/source")    /* Whatever it was originally*/
                     52: 
                     53: </PRE>www/present represents the user's
                     54: perception of the document.  If you
                     55: convert to www/present, you present
                     56: the material to the user. 
                     57: <PRE>#define WWW_PRESENT HTAtom_for("www/present")     /* The user's perception */
                     58: 
                     59: </PRE>The message/rfc822 format means a
                     60: MIME message or a plain text message
                     61: with no MIME header. This is what
                     62: is returned by an HTTP server.
                     63: <PRE>#define WWW_MIME HTAtom_for("www/mime")           /* A MIME message */
                     64: </PRE>www/print is like www/present except
                     65: it represents a printed copy.
                     66: <PRE>#define WWW_PRINT HTAtom_for("www/print") /* A printed copy */
                     67: 
                     68: #define WWW_PLAINTEXT  HTAtom_for("text/plain")
                     69: #define WWW_POSTSCRIPT         HTAtom_for("application/postscript")
                     70: #define WWW_RICHTEXT   HTAtom_for("application/rtf")
                     71: #define WWW_HTML       HTAtom_for("text/html")
                     72: #define WWW_BINARY     HTAtom_for("application/binary")
2.7     ! timbl      73: 
2.1       timbl      74: </PRE>We must include the following file
                     75: after defining HTFormat, to which
                     76: it makes reference.
                     77: <PRE>#include "HTAnchor.h"
                     78: 
                     79: </PRE>
                     80: <H2>The HTPresentation and HTConverter
                     81: types</H2>This HTPresentation structure represents
                     82: a possible conversion algorithm from
                     83: one format to annother.  It includes
                     84: a pointer to a conversion routine.
                     85: The conversion routine returns a
                     86: stream to which data should be fed.
                     87: See also <A
                     88: NAME=z5 HREF="#z3">HTStreamStack</A> which scans
                     89: the list of registered converters
                     90: and calls one. See the <A
                     91: NAME=z6 HREF="HTInit.html">initialisation
                     92: module</A> for a list of conversion routines.
                     93: <PRE>typedef struct _HTPresentation HTPresentation;
                     94: 
2.2       timbl      95: typedef HTStream * HTConverter PARAMS((
2.1       timbl      96:        HTPresentation *        pres,
                     97:        HTParentAnchor *        anchor,
                     98:        HTStream *              sink));
                     99:        
                    100: struct _HTPresentation {
                    101:        HTAtom* rep;            /* representation name atmoized */
                    102:        HTAtom* rep_out;        /* resulting representation */
2.2       timbl     103:        HTConverter *converter; /* The routine to gen the stream stack */
2.1       timbl     104:        char *  command;        /* MIME-format string */
                    105:        float   quality;        /* Between 0 (bad) and 1 (good) */
                    106:        float   secs;
                    107:        float   secs_per_byte;
                    108: };
                    109: 
                    110: </PRE>The list of presentations is kept
                    111: by this module.  It is also scanned
                    112: by modules which want to know the
                    113: set of formats supported. for example.
                    114: <PRE>extern HTList * HTPresentations;
                    115: 
                    116: </PRE>
                    117: <H2>HTSetPresentation: Register a system
                    118: command to present a format</H2>
2.2       timbl     119: <H2>On entry,</H2>
2.1       timbl     120: <DL>
                    121: <DT>rep
                    122: <DD> is the MIME - style format name
                    123: <DT>command
                    124: <DD> is the MAILCAP - style command
                    125: template
                    126: <DT>quality
                    127: <DD> A degradation faction 0..1
                    128: <DT>maxbytes
                    129: <DD> A limit on the length acceptable
                    130: as input (0 infinite)
                    131: <DT>maxsecs
                    132: <DD> A limit on the time user
                    133: will wait (0 for infinity)
                    134: </DL>
                    135: 
                    136: <PRE>extern void HTSetPresentation PARAMS((
                    137:        CONST char * representation,
                    138:        CONST char * command,
                    139:        float   quality,
                    140:        float   secs, 
                    141:        float   secs_per_byte
                    142: ));
                    143: 
                    144: 
                    145: </PRE>
                    146: <H2>HTSetConversion:   Register a converstion
                    147: routine</H2>
2.2       timbl     148: <H2>On entry,</H2>
2.1       timbl     149: <DL>
                    150: <DT>rep_in
                    151: <DD> is the content-type input
                    152: <DT>rep_out
                    153: <DD> is the resulting content-type
                    154: <DT>converter
                    155: <DD> is the routine to make
                    156: the stream to do it
                    157: </DL>
                    158: 
                    159: <PRE>
                    160: extern void HTSetConversion PARAMS((
                    161:        CONST char *    rep_in,
                    162:        CONST char *    rep_out,
2.2       timbl     163:        HTConverter *   converter,
2.1       timbl     164:        float           quality,
                    165:        float           secs, 
                    166:        float           secs_per_byte
                    167: ));
                    168: 
                    169: 
                    170: </PRE>
                    171: <H2><A
                    172: NAME=z3>HTStreamStack:   Create a stack of
                    173: streams</A></H2>This is the routine which actually
                    174: sets up the conversion. It currently
                    175: checks only for direct conversions,
                    176: but indirect conversions are forseen.
2.2       timbl     177: It takes a stream into which the
2.1       timbl     178: output should be sent in the final
                    179: format, builds the conversion stack,
                    180: and returns a stream into which the
                    181: data in the input format should be
                    182: fed.  The anchor is passed because
                    183: hypertxet objects load information
                    184: into the anchor object which represents
                    185: them.
                    186: <PRE>extern HTStream * HTStreamStack PARAMS((
                    187:        HTFormat                format_in,
                    188:        HTFormat                format_out,
                    189:        HTStream*               stream_out,
                    190:        HTParentAnchor*         anchor));
                    191: 
                    192: </PRE>
                    193: <H2>HTStackValue: Find the cost of a
                    194: filter stack</H2>Must return the cost of the same
                    195: stack which HTStreamStack would set
                    196: up.
2.2       timbl     197: <H2>On entry,</H2>
2.1       timbl     198: <DL>
                    199: <DT>format_in
                    200: <DD> The fomat of the data to
                    201: be converted
                    202: <DT>format_out
                    203: <DD> The format required
                    204: <DT>initial_value
                    205: <DD> The intrinsic "value"
                    206: of the data before conversion on
                    207: a scale from 0 to 1
                    208: <DT>length
                    209: <DD> The number of bytes expected
                    210: in the input format
                    211: </DL>
                    212: 
                    213: <PRE>extern float HTStackValue PARAMS((
                    214:        HTFormat                format_in,
                    215:        HTFormat                rep_out,
                    216:        float                   initial_value,
                    217:        long int                length));
                    218: 
                    219: #define NO_VALUE_FOUND -1e20           /* returned if none found */
                    220: 
                    221: </PRE>
                    222: <H2><A
                    223: NAME=z1>HTCopy:  Copy a socket to a stream</A></H2>This is used by the protocol engines
2.6       secret    224: to send data down a stream, typically
2.1       timbl     225: one which has been generated by HTStreamStack.
                    226: <PRE>extern void HTCopy PARAMS((
                    227:        int                     file_number,
                    228:        HTStream*               sink));
                    229: 
                    230:        
2.6       secret    231: </PRE>
                    232: <H2><A
                    233: NAME=c6>HTFileCopy:  Copy a file to a stream</A></H2>This is used by the protocol engines
                    234: to send data down a stream, typically
2.7     ! timbl     235: one which has been generated by HTStreamStack.
        !           236: It is currently called by <A
        !           237: NAME=z9 HREF="#c7">HTParseFile</A>
2.6       secret    238: <PRE>extern void HTFileCopy PARAMS((
                    239:        FILE*                   fp,
                    240:        HTStream*               sink));
                    241: 
                    242:        
2.7     ! timbl     243: </PRE>
        !           244: <H2><A
        !           245: NAME=c2>HTCopyNoCR: Copy a socket to a stream,
        !           246: stripping CR characters.</A></H2>It is slower than <A
2.1       timbl     247: NAME=z2 HREF="#z1">HTCopy</A> .
                    248: <PRE>
                    249: extern void HTCopyNoCR PARAMS((
                    250:        int                     file_number,
                    251:        HTStream*               sink));
                    252: 
                    253: 
                    254: </PRE>
                    255: <H2>Clear input buffer and set file number</H2>This routine and the one below provide
                    256: simple character input from sockets.
                    257: (They are left over from the older
                    258: architecure and may not be used very
                    259: much.)  The existence of a common
                    260: routine and buffer saves memory space
                    261: in small implementations.
                    262: <PRE>extern void HTInitInput PARAMS((int file_number));
                    263: 
                    264: </PRE>
                    265: <H2>Get next character from buffer</H2>
                    266: <PRE>extern char HTGetChararcter NOPARAMS;
                    267: 
                    268: 
                    269: </PRE>
                    270: <H2>HTParseSocket: Parse a socket given
                    271: its format</H2>This routine is called by protocol
                    272: modules to load an object.  uses<A
                    273: NAME=z4 HREF="#z3">
                    274: HTStreamStack</A> and the copy routines
                    275: above.  Returns HT_LOADED if succesful,
                    276: &lt;0 if not.
                    277: <PRE>extern int HTParseSocket PARAMS((
                    278:        HTFormat        format_in,
                    279:        HTFormat        format_out,
                    280:        HTParentAnchor  *anchor,
                    281:        int             file_number,
2.6       secret    282:        HTStream*       sink));
                    283: 
                    284: </PRE>
                    285: <H2><A
2.7     ! timbl     286: NAME=c1>HTParseFile: Parse a File through
        !           287: a file pointer</A></H2>This routine is called by protocols
        !           288: modules to load an object. uses<A
        !           289: NAME=z4 HREF="#z3"> HTStreamStack</A>
        !           290: and <A
        !           291: NAME=c7 HREF="#c6">HTFileCopy</A> .  Returns HT_LOADED
        !           292: if succesful, &lt;0 if not.
2.6       secret    293: <PRE>extern int HTParseFile PARAMS((
                    294:        HTFormat        format_in,
                    295:        HTFormat        format_out,
                    296:        HTParentAnchor  *anchor,
                    297:        FILE            *fp,
2.1       timbl     298:        HTStream*       sink));
                    299: 
                    300: </PRE>
                    301: <H2>Epilogue</H2>
                    302: <PRE>extern BOOL HTOutputSource;       /* Flag: shortcut parser */
                    303: #endif
                    304: 
2.7     ! timbl     305: </PRE>end</A></BODY>

Webmaster