/*
*@@sourcefile textv_html.c:
* this code converts HTML code to escape sequences for the
* XTextView control (textview.c).
*
* This code is in part ugly spaghetti, but this is intentional to
* make this HTML parser FAST. In general, you get about double or
* triple the speed compared to Netscape 4.6 on OS/2. This code
* doesn't understand all of HTML though, but you get most of HTML 2.
* There's no tables or frames at this point.
*
* The entry point into this mess is txvConvertFromHTML, which
* is easy to use.
*
* Note: Version numbering in this file relates to XWorkplace version
* numbering.
*
*@@header "helpers\textv_html.h"
*
*@@added V0.9.3 (2000-05-10) [umoeller]
*/
/*
* Copyright (C) 2000 Ulrich M”ller.
* This program is part of the XWorkplace package.
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, in version 2 as it comes in the COPYING
* file of the XWorkplace main distribution.
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*/
#define OS2EMX_PLAIN_CHAR
// this is needed for "os2emx.h"; if this is defined,
// emx will define PSZ as _signed_ char, otherwise
// as unsigned char
#include
#include
#include
#include
#include "setup.h" // code generation and debugging options
#include "helpers\linklist.h"
#include "helpers\stringh.h"
#include "helpers\textview.h"
#include "helpers\textv_html.h"
/*
*@@category: Helpers\PM helpers\Window classes\XTextView control\HTML conversion
* see textv_html.c.
*/
/* ******************************************************************
*
* Declarations
*
********************************************************************/
/*
*@@ LISTDESC:
* structure stored in COPYTARGET to
* hold list information (UL, OL, ... tags).
*
*@@added V0.9.3 (2000-05-07) [umoeller]
*/
typedef struct _LISTDESC
{
ULONG ulListType; // 0: unordered (UL)
// 1: ordered (OL)
// 2: definition lists (DL)
ULONG ulItem; // list enumeration; 1 on first item,
// 2 on next, ...
} LISTDESC, *PLISTDESC;
/*
*@@ COPYTARGET:
* monster structure which holds the current
* status of the HTML converter while conversion
* is taking place. This stores input/output pointers
* and various flags to avoid duplicate line breaks
* and such.
*
* One instance of this is created in txvConvertFromHTML
* on the stack and then passed to all the sub-function
* calls.
*
*@@added V0.9.3 (2000-05-06) [umoeller]
*/
typedef struct _COPYTARGET
{
PSZ pSource; // ptr into source string;
// valid ONLY while we're in a tag handler
PSZ pNewSource; // can be set by tag handler to skip characters;
// this is set to NULL before calling a tag
// handler; if this is still NULL, default
// processing occurs
// new string:
PSZ pszNew; // memory buffer
ULONG cbNew; // size of buffer (reallocated)
PSZ pTarget; // current char ptr into pszNew
// saved character while tag handler is being called
CHAR cSaved;
PSZ *ppszTitle; // out: title (ptr can be NULL)
// V0.9.20 (2002-08-10) [umoeller]
// formatting flags while going through the text
BOOL fSkipNextSpace;
// if TRUE, subsequent spaces are skipped
BOOL fNeedsLinebreak;
// if TRUE, \n is inserted before any other character
BOOL fSkipNextLinebreak;
// if TRUE, subsequent linebreaks are skipped
BOOL fPRE;
// are we currently in a PRE tag?
BOOL fInLink;
// are we currently in a A HREF= tag?
// arguments (attributes) for tag handlers
PSZ pszAttributes; // != NULL while a tag handler is being called
// and attributes exist for the tag
// anchors count
// USHORT usAnchorIndex; // start with 1 removed V0.9.20 (2002-08-10) [umoeller]
// list maintenance
ULONG ulListLevel; // if > 0, we're in a UL or OL block;
// raised for each block
ULONG ulUnorderedListLevel; // raised with each UL block to keep track
// of bullets
ULONG ulOrderedListLevel; // raised with each UL block to keep track
// of 1), 2), a), b)... numbering
ULONG ulCurrentListType; // current list type (from highest LISTDESC)
BOOL fInDT; // TRUE if we're currently in a DT tag
LINKLIST llLists; // stack of LISTDESC items
} COPYTARGET, *PCOPYTARGET;
typedef VOID FNPROCESSTAG(PCOPYTARGET pct);
typedef FNPROCESSTAG *PFNPROCESSTAG;
/* ******************************************************************
*
* Global variables
*
********************************************************************/
/* ******************************************************************
*
* Append-char helpers
*
********************************************************************/
#define COPYTARGETALLOC 100000
/*
*@@ AppendChar:
* helper for txvConvertFromHTML to
* append a char to the target string
* in COPYTARGET.
* This performs a few additional checks
* and manages memory.
*
*@@added V0.9.3 (2000-05-06) [umoeller]
*/
STATIC VOID AppendChar(PCOPYTARGET pct, // in/out: formatting buffer
unsigned char c)
{
// calculate ofs where to store next char
ULONG cbOfsNext = pct->pTarget - pct->pszNew;
if (cbOfsNext >= pct->cbNew) // have we reached the buffer size yet?
{
// more mem needed:
pct->cbNew += COPYTARGETALLOC;
pct->pszNew = (PSZ)realloc(pct->pszNew, pct->cbNew);
// if first call, pszNew is NULL, and realloc
// behaves just like malloc
// adjust target, because ptr might have changed
pct->pTarget = pct->pszNew + cbOfsNext;
}
// append character
*pct->pTarget++ = c;
}
/*
*@@ AppendString:
* appends the characters in *ach,
* which must be null-terminated.
* Does NOT append a null character though.
*
*@@added V0.9.3 (2000-05-06) [umoeller]
*/
STATIC VOID AppendString(PCOPYTARGET pct, // in/out: formatting buffer
char *ach)
{
ULONG cbAppend = strlen(ach);
ULONG ul;
PSZ pSource;
// calculate ofs where to store next char
ULONG cbOfsNext = pct->pTarget - pct->pszNew;
while (cbOfsNext + cbAppend >= pct->cbNew)
{
// more mem needed:
pct->cbNew += COPYTARGETALLOC;
pct->pszNew = (PSZ)realloc(pct->pszNew, pct->cbNew);
// if first call, pszNew is NULL, and realloc
// behaves just like malloc
// adjust target, because ptr might have changed
pct->pTarget = pct->pszNew + cbOfsNext;
}
// append characters
pSource = ach;
for (ul = 0;
ul < cbAppend;
ul++)
*pct->pTarget++ = *pSource++;
}
/*
*@@ AppendLinebreakCheck:
* checks if a linebreak is needed and
* inserts one if so.
*
*@@added V0.9.3 (2000-05-17) [umoeller]
*/
STATIC VOID AppendLinebreakCheck(PCOPYTARGET pct)
{
if ((!pct->fPRE) && (pct->fNeedsLinebreak))
{
// yes: insert linebreak; this resets pct->fNeedsLinebreak
if (!pct->fSkipNextLinebreak)
{
AppendChar(pct, '\n');
if ((pct->ulListLevel) && (!pct->fInDT))
// if we're in a list, add a tab also,
// because we'll have a negative first-line margin
AppendString(pct, TXVESC_TAB);
}
pct->fNeedsLinebreak = FALSE;
}
}
/*
*@@ AppendEscapeWithDecimal:
* appends the specified escape code
* with a three-digit decimal parameter.
* Calls AppendString in turn.
*
*@@added V0.9.3 (2000-05-07) [umoeller]
*/
STATIC VOID AppendEscapeWith3Decimals(PCOPYTARGET pct, // in/out: formatting buffer
char *ach,
USHORT us)
{
CHAR szDecimal[10];
if (us > 999)
us = 999;
sprintf(szDecimal, "%03d", us);
// append escape
AppendString(pct, ach);
AppendString(pct, szDecimal);
}
/*
*@@ AppendEscapeWith4Decimals:
*
*@@added V0.9.3 (2000-05-07) [umoeller]
*/
STATIC VOID AppendEscapeWith4Decimals(PCOPYTARGET pct, // in/out: formatting buffer
char *ach,
USHORT us)
{
CHAR szDecimal[10];
if (us > 9999)
us = 9999;
sprintf(szDecimal, "%04d", us);
// append escape
AppendString(pct, ach);
AppendString(pct, szDecimal);
}
/* ******************************************************************
*
* Tag converter functions
*
********************************************************************/
/*
*@@ StartList:
* starts a list (UL or OL).
* This uses a linked list in COPYTARGET
* to keep a pseudo-stack for nested lists.
*
*@@added V0.9.3 (2000-05-08) [umoeller]
*/
STATIC VOID StartList(PCOPYTARGET pct, // in/out: formatting buffer
ULONG ulListType) // list type:
// 0: unordered (UL)
// 1: ordered (OL)
// 2: definition lists (DL)
{
PLISTDESC pListDesc;
// raise list level
pct->ulListLevel++;
if (ulListType == 0)
// unordered:
pct->ulUnorderedListLevel++;
else if (ulListType == 1)
// ordered:
pct->ulOrderedListLevel++;
// create LISTDESC and store it on stack
pListDesc = (PLISTDESC)malloc(sizeof(LISTDESC));
pListDesc->ulListType
= pct->ulCurrentListType
= ulListType;
pListDesc->ulItem = 1;
lstAppendItem(&pct->llLists,
pListDesc);
AppendEscapeWith4Decimals(pct,
TXVESC_LEFTMARGIN,
pct->ulListLevel * 5);
AppendEscapeWith3Decimals(pct,
TXVESC_FIRSTLINEMARGIN_LEFT,
(ulListType == 2)
? 5 // for definition lists
: 3); // negative!
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
/*
*@@ StopList:
* stops a list (UL or OL).
*
*@@added V0.9.3 (2000-05-07) [umoeller]
*/
STATIC VOID StopList(PCOPYTARGET pct)
{
if (pct->ulListLevel)
{
PLISTNODE pNode;
// lower list level
pct->ulListLevel--;
AppendEscapeWith4Decimals(pct,
TXVESC_LEFTMARGIN,
pct->ulListLevel * 5);
AppendEscapeWith3Decimals(pct,
TXVESC_FIRSTLINEMARGIN_LEFT,
(pct->ulListLevel)
? 3 // we still have a list level (nested)
: 0);
pct->fNeedsLinebreak = TRUE;
// remove the LISTDESC from the stack
pNode = lstNodeFromIndex(&pct->llLists,
pct->ulListLevel); // this has been lowered already
if (pNode)
{
PLISTDESC pListDesc = (PLISTDESC)pNode->pItemData;
if (pListDesc->ulListType == 0)
// was unordered:
pct->ulUnorderedListLevel--;
else if (pListDesc->ulListType == 1)
// was ordered:
pct->ulOrderedListLevel--;
lstRemoveNode(&pct->llLists, pNode);
// update COPYTARGET with previous list level
if (pct->ulListLevel)
{
// we're still in a list (nested lists):
PLISTDESC pListDesc2 = (PLISTDESC)lstItemFromIndex(&pct->llLists,
pct->ulListLevel - 1);
if (pListDesc2)
pct->ulCurrentListType = pListDesc2->ulListType;
}
}
}
// else: buggy HTML code, ignore
}
/*
*@@ TagTITLE:
*
*@@added V0.9.3 (2000-05-19) [umoeller]
*/
STATIC VOID TagTITLE(PCOPYTARGET pct)
{
// pSource currently points to pSource + strlen(pct->pSource);
// points to temporary null byte in main buffer now
*pSource = pct->cSaved;
if (pSource = strchr(pct->pSource, '>'))
{
PSZ pNextOpen;
if (pNextOpen = strchr(pSource, '<'))
{
// extract title
if (pct->ppszTitle)
*(pct->ppszTitle) = strhSubstr(pSource + 1, pNextOpen);
// adjusted V0.9.20 (2002-08-10) [umoeller]
if (strnicmp(pNextOpen + 1, "/TITLE", 6) == 0)
{
// closing /TITLE tag found:
// search on after that
if (pct->pNewSource = strchr(pNextOpen, '>'))
pct->pNewSource++;
}
}
}
}
/*
*@@ TagP:
*
*/
STATIC VOID TagP(PCOPYTARGET pct)
{
// append newline:
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
/* if (pct->ulListLevel)
{
// if we are currently in a list, we must also
// add a tab escape, because we have set
// the first line margin to the left of the
// left margin
AppendString(pct,
TXVESC_TAB);
} */
}
STATIC VOID TagBR(PCOPYTARGET pct)
{
AppendChar(pct,
'\r');
if (pct->ulListLevel)
{
// if we are currently in a list, we must also
// add a tab escape, because we have set
// the first line margin to the left of the
// left margin
AppendString(pct,
TXVESC_TAB);
}
if (!pct->fPRE)
pct->fSkipNextSpace = TRUE;
}
STATIC VOID TagPRE(PCOPYTARGET pct)
{
// start of PRE tag:
// add \n before any other character
// pct->fNeedsLinebreak = TRUE;
AppendChar(pct, '\n');
pct->fNeedsLinebreak = FALSE;
/* AppendString(pct,
TXVESC_PRE_BEGIN); */
AppendEscapeWith3Decimals(pct,
TXVESC_SET_FONT,
1); // monospaced font
AppendEscapeWith4Decimals(pct,
TXVESC_SPACEBEFORE,
0); // no spacing before
AppendEscapeWith4Decimals(pct,
TXVESC_SPACEAFTER,
0); // no spacing after
// disable word-wrapping
AppendString(pct,
TXVESC_WORDWRAP "0");
pct->fPRE = TRUE;
pct->fSkipNextSpace = FALSE;
}
STATIC VOID TagXPRE(PCOPYTARGET pct)
{
pct->fPRE = FALSE;
AppendEscapeWith3Decimals(pct,
TXVESC_SET_FONT,
0); // standard font
AppendString(pct, TXVESC_SPACEBEFORE);
AppendString(pct, "####"); // reset to default
AppendString(pct, TXVESC_SPACEAFTER);
AppendString(pct, "####"); // reset to default
// re-enable word-wrapping
AppendString(pct,
TXVESC_WORDWRAP "1"
"\n"); // force line break
pct->fNeedsLinebreak = FALSE;
// refuse to add \n even if we have another "p" coming up
pct->fSkipNextLinebreak = TRUE;
pct->fSkipNextSpace = TRUE;
}
STATIC VOID TagH1(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
200); // double size
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXH1(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // regular size
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagH2(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
175); // size in percent of regular point size
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXH2(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // regular size
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagH3(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
150); // size in percent of regular point size
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXH3(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // size in percent of regular point size
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagH4(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
125); // size in percent of regular point size
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXH4(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // regular size
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagH5(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // size in percent of regular point size
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXH5(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // regular size
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagH6(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
80 ); // size in percent of regular point size
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXH6(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
AppendEscapeWith3Decimals(pct,
TXVESC_POINTSIZE_REL,
100); // regular size
// add \n before any other character
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagUL(PCOPYTARGET pct)
{
StartList(pct,
0); // unordered
}
STATIC VOID TagXUL(PCOPYTARGET pct)
{
StopList(pct);
}
STATIC VOID TagOL(PCOPYTARGET pct)
{
StartList(pct,
1); // ordered
}
STATIC VOID TagXOL(PCOPYTARGET pct)
{
StopList(pct);
}
STATIC VOID TagLI(PCOPYTARGET pct)
{
PLISTDESC pListDesc;
CHAR szMarker[20] = TXVESC_MARKER "\x01";
if (pct->ulListLevel)
{
// we're in a list:
pListDesc = (PLISTDESC)lstItemFromIndex(&pct->llLists,
pct->ulListLevel - 1);
if (pListDesc)
{
if (pListDesc->ulListType == 1)
// is ordered list:
sprintf(szMarker, "%lu.", (pListDesc->ulItem)++);
else if (pListDesc->ulListType == 0)
// is unordered list:
// set bullet type according to unordered nesting
szMarker[2] = pct->ulUnorderedListLevel;
}
}
// add \n before any other character
// pct->fNeedsLinebreak = TRUE;
// if (pct->fNeedsLinebreak)
{
AppendChar(pct, '\n');
pct->fNeedsLinebreak = FALSE;
}
AppendString(pct, szMarker);
AppendString(pct, TXVESC_TAB);
}
STATIC VOID TagDL(PCOPYTARGET pct)
{
StartList(pct,
2); // definition list
}
STATIC VOID TagXDL(PCOPYTARGET pct)
{
StopList(pct);
pct->fInDT = FALSE;
}
STATIC VOID TagDT(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
pct->fInDT = TRUE;
}
STATIC VOID TagDD(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
AppendString(pct, TXVESC_TAB);
if (!pct->fPRE)
pct->fSkipNextSpace = TRUE;
pct->fInDT = FALSE;
}
STATIC VOID TagTR(PCOPYTARGET pct)
{
pct->fNeedsLinebreak = TRUE;
}
STATIC VOID TagB(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_BEGIN);
}
STATIC VOID TagXB(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_BOLD_END);
}
STATIC VOID TagI(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_ITALICS_BEGIN);
}
STATIC VOID TagXI(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_ITALICS_END);
}
STATIC VOID TagU(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_UNDERLINE_BEGIN);
}
STATIC VOID TagXU(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_UNDERLINE_END);
}
STATIC VOID TagSTRIKE(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_STRIKE_BEGIN);
}
STATIC VOID TagXSTRIKE(PCOPYTARGET pct)
{
AppendString(pct,
TXVESC_STRIKE_END);
}
STATIC VOID TagCODE(PCOPYTARGET pct)
{
AppendEscapeWith3Decimals(pct,
TXVESC_SET_FONT,
1); // monospaced font
}
STATIC VOID TagXCODE(PCOPYTARGET pct)
{
AppendEscapeWith3Decimals(pct,
TXVESC_SET_FONT,
0); // regular font
}
STATIC VOID TagA(PCOPYTARGET pct)
{
CHAR szAnchor[10];
PSZ pHREF = NULL;
pct->fInLink = FALSE;
if (pct->pszAttributes)
{
// we have attributes:
PSZ pszClosingTag;
if (pszClosingTag = strchr(pct->pszAttributes, '>'))
{
ULONG ulOfs = 0;
/*
* HREF attribute:
*
*/
PSZ pNAME = 0;
// replace '>' with null char to mark end of search
*pszClosingTag = 0;
if (pHREF = strhGetTextAttr(pct->pszAttributes, "HREF", &ulOfs))
// OK, we got a link target:
pct->fInLink = TRUE;
// do not free
/*
* NAME attribute:
*
*/
if (pNAME = strhGetTextAttr(pct->pszAttributes, "NAME", &ulOfs))
{
AppendString(pct,
TXVESC_ANCHORNAME);
AppendString(pct,
pNAME);
// must be terminated with 0xFF
AppendChar(pct, 0xFF);
free(pNAME);
}
// restore '>'
*pszClosingTag = '>';
}
}
if (pHREF)
{
AppendString(pct,
TXVESC_LINK_BEGIN);
AppendString(pct,
pHREF);
// must be terminated with 0xFF
AppendChar(pct, 0xFF);
free(pHREF);
}
}
STATIC VOID TagXA(PCOPYTARGET pct)
{
if (pct->fInLink)
{
AppendString(pct,
TXVESC_LINK_END);
pct->fInLink = FALSE;
}
}
/* ******************************************************************
*
* Tag helpers
*
********************************************************************/
/*
*@@ FindTagProcessor:
* returns the Tag* function which handles the
* given tag or NULL if there's none.
*
*@@added V0.9.4 (2000-06-10) [umoeller]
*/
STATIC PFNPROCESSTAG FindTagProcessor(PSZ pszTag)
{
PFNPROCESSTAG pProcessor = NULL;
CHAR c0,
c1;
BOOL fEndOfTag = FALSE;
PSZ pCheck = pszTag,
p2;
if (*pCheck == '/')
{
// end of tag:
fEndOfTag = TRUE;
pCheck++;
}
c0 = *pCheck;
c1 = *(pCheck + 1);
p2 = pCheck + 2;
switch (c0)
{
case 'A':
case 'a':
switch (c1)
{
case 0: // A
if (!fEndOfTag)
return TagA;
else
return TagXA;
case 'D': // ADDRESS
case 'd': // ADDRESS
if (stricmp(p2, "DRESS") == 0)
{
if (!fEndOfTag)
return TagI;
else
return TagXI;
}
}
break;
case 'B':
case 'b':
switch (c1)
{
case 0:
if (!fEndOfTag)
return TagB;
else
return TagXB;
case 'R': // BR
case 'r': // BR
if (*p2 == 0)
if (!fEndOfTag)
return TagBR;
}
break;
case 'C':
case 'c':
switch (c1)
{
case 'I': // CITE
case 'i': // CITE
if (stricmp(p2, "TE") == 0)
{
if (!fEndOfTag)
return TagI;
else
return TagXI;
}
break;
case 'O':
case 'o':
if (stricmp(p2, "DE") == 0)
{
if (!fEndOfTag)
return TagCODE;
else
return TagXCODE;
}
break;
}
break;
case 'D':
case 'd':
switch (c1)
{
case 'D': // DD
case 'd': // DD
if ((*p2 == 0) && (!fEndOfTag))
return TagDD;
break;
case 'I': // DIR
case 'i': // DIR
if (*p2 == 'R')
if (*(pCheck + 3) == 0)
{
if (!fEndOfTag)
return TagUL;
else
return TagXUL;
}
break;
case 'L': // DL
case 'l': // DL
if (*p2 == 0)
{
if (!fEndOfTag)
return TagDL;
else
return TagXDL;
}
break;
case 'T': // DT
case 't': // DT
if ((*p2 == 0) && (!fEndOfTag))
return TagDT;
break;
}
break;
case 'E':
case 'e':
if ( (c1 == 'M') || (c1 == 'm') ) // EM
if (*p2 == 0)
{
if (!fEndOfTag)
return TagI;
else
return TagXI;
}
break;
case 'H':
case 'h':
if (c1)
if (*p2 == 0)
switch (c1)
{
case '1':
if (!fEndOfTag)
return TagH1;
else
return TagXH1;
case '2':
if (!fEndOfTag)
return TagH2;
else
return TagXH2;
case '3':
if (!fEndOfTag)
return TagH3;
else
return TagXH3;
case '4':
if (!fEndOfTag)
return TagH4;
else
return TagXH4;
case '5':
if (!fEndOfTag)
return TagH5;
else
return TagXH5;
case '6':
if (!fEndOfTag)
return TagH6;
else
return TagXH6;
}
break;
case 'I':
case 'i':
if (c1 == 0)
{
if (!fEndOfTag)
return TagI;
else
return TagXI;
}
break;
case 'L':
case 'l':
if ((c1 == 'I') || (c1 == 'i'))
if (*p2 == 0)
return TagLI;
break;
case 'M':
case 'm':
if (stricmp(p2, "NU") == 0)
{
if (!fEndOfTag)
return TagUL;
else
return TagXUL;
}
break;
case 'O':
case 'o':
if ((c1 == 'L') || (c1 == 'l'))
if (*p2 == 0)
{
if (!fEndOfTag)
return TagOL;
else
return TagXOL;
}
break;
case 'P':
case 'p':
switch (c1)
{
case 0:
if (!fEndOfTag)
return TagP;
break;
case 'R': // PRE
case 'r': // PRE
if ((*p2 == 'E') || (*p2 == 'e'))
if (*(pCheck + 3) == 0)
{
if (!fEndOfTag)
return TagPRE;
else
return TagXPRE;
}
break;
}
break;
case 'S':
case 's':
switch (c1)
{
case 'T': // STRONG
case 't': // STRONG
if (stricmp(p2, "RONG") == 0)
{
if (!fEndOfTag)
return TagB;
else
return TagXB;
}
else if (stricmp(p2, "RIKE") == 0)
{
if (!fEndOfTag)
return TagSTRIKE;
else
return TagXSTRIKE;
}
break;
case 'A':
case 'a':
if (stricmp(p2, "MP") == 0)
{
if (!fEndOfTag)
return TagCODE;
else
return TagXCODE;
}
break;
}
break;
case 'T':
case 't':
switch (c1)
{
case 'R':
case 'r':
if (*p2 == 0)
return TagTR;
break;
case 'I':
case 'i':
if (stricmp(p2, "TLE") == 0)
return TagTITLE;
break;
case 'T': // TT
case 't':
if (*p2 == 0)
{
if (!fEndOfTag)
return TagCODE;
else
return TagXCODE;
}
break;
}
break;
case 'U':
case 'u':
switch (c1)
{
case 0:
if (!fEndOfTag)
return TagU;
else
return TagXU;
case 'L':
case 'l':
if (*p2 == 0)
{
if (!fEndOfTag)
return TagUL;
else
return TagXUL;
}
break;
}
break;
case 'V':
case 'v':
if (stricmp(p2, "R") == 0)
{
if (!fEndOfTag)
return TagI;
else
return TagXI;
}
break;
case 'X':
case 'x':
if (stricmp(p2, "MP") == 0) // XMP
{
if (!fEndOfTag)
return TagPRE;
else
return TagXPRE;
}
break;
}
return pProcessor;
}
/*
*@@ HandleTag:
* called by txvConvertFromHTML when a "<" character
* is found in the source buffer. This calls
* FindTagProcessor in turn to find the Tag*
* function which handles the tag.
*
*@@added V0.9.3 (2000-05-18) [umoeller]
*/
STATIC VOID HandleTag(PCOPYTARGET pct)
{
PSZ pStartOfTag = pct->pSource;
// '<' == begin of tag:
// is it a comment? <!-- ... -->
if (strncmp(pStartOfTag + 1, "!--", 3) == 0)
{
// start of comment:
// find end of comment
PSZ pEnd = strstr(pStartOfTag, "-->");
if (pEnd)
// found:
// search on after end of comment
pct->pSource = pEnd + 3;
else
{
// end of comment not found:
// stop formatting...
pct->pSource++;
return;
}
}
else
{
// no comment:
// find end of tag
PSZ p2 = pStartOfTag + 1,
pNextClose = 0, // receives first '>' after '<'
pNextSpace = 0; // receives first ' ' after '<'
BOOL fCont = TRUE;
while (fCont)
{
switch (*p2)
{
case ' ':
case '\r':
case '\n':
// store first space after '<'
if (!pNextSpace)
pNextSpace = p2;
// overwrite line breaks with spaces;
// otherwise we cannot handle tags which go across
// several lines, which is valid HTML
*p2 = ' ';
break;
case '>': // end of tag found:
pNextClose = p2;
fCont = FALSE;
break;
case '<':
// another opening tag:
// that's an HTML error
AppendChar(pct,
*pct->pSource++);
fCont = FALSE;
break;
case 0:
fCont = FALSE;
break;
}
p2++;
}
if (pNextClose)
{
// end of tag found:
ULONG cbTag;
// PSZ pStartOfAttrs = 0;
if ((pNextSpace) && (pNextSpace < pNextClose))
{
// we have attributes:
cbTag = pNextSpace - (pStartOfTag + 1);
// pStartOfAttrs = pNextSpace;
}
else
cbTag = pNextClose - (pStartOfTag + 1);
if (!cbTag)
{
// happens if we have a "<>" in the text:
// just insert the '<>' and go on, we have no tag here
AppendChar(pct,
*pct->pSource++);
AppendChar(pct,
*pct->pSource++);
}
else
{
PFNPROCESSTAG pTagProcessor;
pct->cSaved = *(pStartOfTag + cbTag + 1);
// add a null terminator
*(pStartOfTag + cbTag + 1) = 0;
// find corresponding tag converter function
// from G_TagProcessors map
pTagProcessor = FindTagProcessor(pStartOfTag + 1); // pszTag);
// restore char under null terminator
*(pStartOfTag + cbTag + 1) = pct->cSaved;
// reset new source ptr; the tag handler
// can modify this
pct->pNewSource = NULL;
if (pTagProcessor)
{
// tag understood:
// terminate string after closing tag
pct->cSaved = *(pNextClose + 1); // can be null byte!
*(pNextClose + 1) = 0;
// did we have attributes?
if (pNextSpace)
pct->pszAttributes = pNextSpace;
// finally, call the tag handler
(pTagProcessor) // function
(pct); // argument
*(pNextClose + 1) = pct->cSaved;
}
if (pct->pNewSource == NULL)
// tag handler needs no special processing:
// skip '>' too
pct->pSource = pNextClose + 1;
else
// tag handler has skipped something:
pct->pSource = pct->pNewSource;
}
}
}
}
/*
*@@ ConvertEscape:
* called by HandleEscape to find the ANSI (CP 1004)
* character for the given escape sequence (pszTag).
*
* pszTag must be null-terminated and contain only
* the stuff between "&" and ";".
*
* This is really ugly spaghetti, but it's the fastest
* way to do it.
*
*@@added V0.9.4 (2000-06-10) [umoeller]
*/
STATIC unsigned char ConvertEscape(PSZ pszTag)
{
CHAR c0, c1;
CHAR crc = 0;
PSZ p2 = pszTag + 2;
c0 = *pszTag;
c1 = *(pszTag + 1);
switch (c0)
{
case 'a':
switch (c1)
{
case 'a':
if (strcmp(p2, "cute") == 0)
return 225;
break;
case 'c':
if (strcmp(p2, "irc") == 0)
return 226;
else if (strcmp(p2, "ute") == 0)
return 180;
break;
case 'e':
if (strcmp(p2, "lig") == 0)
return 230;
break;
case 'g':
if (strcmp(p2, "rave") == 0)
return 224;
break;
case 'm':
if (strcmp(p2, "p") == 0)
return '&';
break;
case 'r':
if (strcmp(p2, "ing") == 0)
return 229;
break;
case 't':
if (strcmp(p2, "ilde") == 0)
return 227;
break;
case 'u':
if (strcmp(p2, "ml") == 0)
return 228;
break;
}
break;
case 'b':
if (strcmp(pszTag + 1, "rvbar") == 0)
return 166;
break;
case 'c':
switch (c1)
{
case 'c':
if (strcmp(p2, "edil") == 0)
return 231;
break;
case 'e':
if (strcmp(p2, "dil") == 0)
return 184;
else if (strcmp(p2, "nt") == 0)
return 162;
break;
case 'o':
if (strcmp(p2, "py") == 0)
return 169;
break;
case 'u':
if (strcmp(p2, "rren") == 0)
return 164;
}
break;
case 'd':
switch (c1)
{
case 'e':
if (strcmp(p2, "g") == 0) return 176;
break;
case 'i':
if (strcmp(p2, "vide") == 0) return 247;
break;
}
break;
case 'e':
switch (c1)
{
case 'a':
if (strcmp(p2, "cute") == 0) return 233;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 234;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 232;
break;
case 't':
if (strcmp(p2, "h") == 0) return 240;
break;
case 'u':
if (strcmp(p2, "ml") == 0) return 235;
break;
}
break;
case 'f':
switch (c1)
{
case 'r':
if (strcmp(p2, "ac14") == 0) return 188;
if (strcmp(p2, "ac12") == 0) return 189;
if (strcmp(p2, "ac34") == 0) return 190;
break;
}
break;
case 'g':
switch (c1)
{
case 't':
if (*p2 == 0) return '>';
}
break;
case 'i':
switch (c1)
{
case 'a':
if (strcmp(p2, "cute") == 0) return 237;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 238;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 236;
break;
case 'e':
if (strcmp(p2, "xcl") == 0) return 161;
break;
case 'q':
if (strcmp(p2, "uest") == 0) return 191;
break;
case 'u':
if (strcmp(p2, "ml") == 0) return 239;
}
break;
case 'l':
switch (c1)
{
case 't':
if (*p2 == 0)
return '<';
break;
case 'a':
if (strcmp(p2, "quo") == 0) return 171;
}
break;
case 'm':
switch (c1)
{
case 'a':
if (strcmp(p2, "cr") == 0) return 175;
break;
case 'i':
if (strcmp(p2, "cro") == 0) return 181;
if (strcmp(p2, "ddot") == 0) return 183;
break;
}
break;
case 'n':
switch (c1)
{
case 'b':
if (strcmp(p2, "sp") == 0) return 160;
break;
case 'o':
if (strcmp(p2, "t") == 0) return 172;
break;
case 't':
if (strcmp(p2, "ilde") == 0) return 241;
}
break;
case 'o':
switch (c1)
{
case 'a':
if (strcmp(p2, "cute") == 0) return 243;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 244;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 242;
break;
case 'r':
if (strcmp(p2, "df") == 0) return 170;
if (strcmp(p2, "dm") == 0) return 186;
break;
case 's':
if (strcmp(p2, "lash") == 0) return 248;
break;
case 't':
if (strcmp(p2, "ilde") == 0) return 245;
break;
case 'u':
if (strcmp(p2, "ml") == 0) return 246;
}
break;
case 'p':
switch (c1)
{
case 'a':
if (strcmp(p2, "ra") == 0) return 182;
break;
case 'l':
if (strcmp(p2, "usmn") == 0) return 177;
break;
case 'o':
if (strcmp(p2, "und") == 0) return 163;
}
break;
case 'q':
if (strcmp(pszTag, "quot") == 0) return '"';
break;
case 'r':
if (strcmp(pszTag, "raquo") == 0) return 187;
if (strcmp(pszTag, "reg") == 0) return 174;
break;
case 's':
switch (c1)
{
case 'z':
if (strcmp(p2, "lig") == 0) return 223;
break;
case 'e':
if (strcmp(p2, "ct") == 0) return 167;
break;
case 'h':
if (strcmp(p2, "y") == 0) return 173;
break;
case 'u':
if (strcmp(p2, "p1") == 0) return 185;
if (strcmp(p2, "p2") == 0) return 178;
if (strcmp(p2, "p3") == 0) return 179;
}
break;
case 't':
if (strcmp(pszTag, "thorn") == 0) return 254;
if (strcmp(pszTag, "times") == 0) return 215;
break;
case 'u':
switch (c1)
{
case 'a':
if (strcmp(p2, "cute") == 0) return 250;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 251;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 249;
break;
case 'm':
if (strcmp(p2, "l") == 0) return 168;
break;
case 'u':
if (strcmp(p2, "ml") == 0) return 252;
}
break;
case 'y':
if (strcmp(pszTag, "yacute") == 0) return 253;
if (strcmp(pszTag, "yen") == 0) return 165;
if (strcmp(pszTag, "yuml") == 0) return 255;
break;
case 'A':
switch (c1)
{
case 'u':
if (strcmp(p2, "ml") == 0) return 196;
break;
case 'a':
if (strcmp(p2, "cute") == 0) return 193;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 194;
break;
case 'E':
if (strcmp(p2, "lig") == 0) return 198;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 192;
break;
case 'r':
if (strcmp(p2, "ing") == 0) return 197;
break;
case 't':
if (strcmp(p2, "ilde") == 0) return 195;
}
break;
case 'C':
if (strcmp(pszTag, "Ccedil") == 0) return 199;
break;
case 'E':
if (strcmp(pszTag, "Ecirc") == 0) return 202;
if (strcmp(pszTag, "Eacute") == 0) return 201;
if (strcmp(pszTag, "Egrave") == 0) return 200;
if (strcmp(pszTag, "ETH") == 0) return 208;
if (strcmp(pszTag, "Euml") == 0) return 203;
break;
case 'I':
if (strcmp(pszTag, "Icirc") == 0) return 206;
if (strcmp(pszTag, "Iacute") == 0) return 205;
if (strcmp(pszTag, "Igrave") == 0) return 204;
if (strcmp(pszTag, "Iuml") == 0) return 207;
break;
case 'N':
if (strcmp(pszTag, "Ntilde") == 0) return 209;
break;
case 'O':
switch (c1)
{
case 'u':
if (strcmp(p2, "ml") == 0) return 214;
break;
case 'a':
if (strcmp(p2, "cute") == 0) return 211;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 212;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 210;
break;
case 't':
if (strcmp(p2, "ilde") == 0) return 213;
break;
case 's':
if (strcmp(p2, "lash") == 0) return 216;
}
break;
case 'U':
switch (c1)
{
case 'a':
if (strcmp(p2, "cute") == 0) return 218;
break;
case 'c':
if (strcmp(p2, "irc") == 0) return 219;
break;
case 'g':
if (strcmp(p2, "rave") == 0) return 217;
break;
case 'u':
if (strcmp(p2, "ml") == 0) return 220;
}
break;
case 'T':
if (strcmp(pszTag, "THORN") == 0) return 222;
break;
case 'Y':
if (strcmp(pszTag, "Yacute") == 0) return 221;
break;
}
return crc;
}
/*
*@@ HandleEscape:
* called by txvConvertFromHTML when a "&" character
* is found in the source buffer. This calls
* ConvertEscape in turn.
*
*@@added V0.9.3 (2000-05-18) [umoeller]
*/
STATIC VOID HandleEscape(PCOPYTARGET pct)
{
// ampersand:
// replace special characters
PSZ pStartOfTag = pct->pSource;
// find end of tag
PSZ p2 = pStartOfTag,
pNextClose = 0,
pNextSpace = 0;
BOOL fCont = TRUE;
while (fCont)
{
switch (*p2)
{
case 0:
fCont = FALSE;
break;
case ';':
pNextClose = p2;
fCont = FALSE;
break;
case ' ':
if (!pNextSpace)
pNextSpace = p2;
break;
}
p2++;
}
if (!pNextClose)
// no closing tag found:
// just insert the '&' and go on, we have no tag here
AppendChar(pct,
*pct->pSource++);
else
{
if ((pNextSpace) && (pNextSpace < pNextClose))
// space before ';':
// just insert the '&' and go on, we have no tag here
AppendChar(pct,
*pct->pSource++);
else if ((!pNextClose) || (pNextClose <= pStartOfTag + 1))
AppendChar(pct,
*pct->pSource++);
else
{
ULONG ulCode = 0;
// create substring with tag
PSZ pszTag = pStartOfTag + 1;
*pNextClose = 0;
if (*pszTag == '#')
{
// latin-1 or Unicode encoding (�)
ulCode = atoi(pszTag + 1);
// next input: char after ';'
pct->pSource = pNextClose + 1;
}
else
{
// named entity:
// find char code corresponding to escape
// from G_EscapeProcessors map
ulCode = ConvertEscape(pszTag);
if (ulCode)
// tag supported:
pct->pSource = pNextClose + 1;
else
// tag not supported:
ulCode = *pct->pSource++;
}
// restore closing tag which we overwrote
*pNextClose = ';';
if (ulCode)
{
AppendLinebreakCheck(pct);
AppendChar(pct,
(CHAR)ulCode);
pct->fSkipNextSpace = FALSE;
}
}
}
}
/* ******************************************************************
*
* Entry points
*
********************************************************************/
/*
*@@ txvConvertFromHTML:
* this modifies the given text string (which should
* be the complete BODY block of any HTML file) so
* that all HTML tags are removed and replaced with
* escape sequences that the XTextView control understands.
*
* The buffer gets reallocated by this function, so it
* must be free()'able.
*
* So, to have the XTextView control display an HTML file,
* do this:
*
* 1) Load an HTML file into a buffer allocated by malloc().
*
* 2) Call txvConvertFromHTML.
*
* 3) Call WinSetWindowText on the XTextView control with
* the modified buffer.
*
* This understands the following limited subset of HTML:
*
* Paragraph tags:
*
* -- P, BR
* -- PRE, /PRE
* -- UL, /UL, OL, /OL, LI
* -- DL, /DL, DT, DD
* -- H1, /H1 thru H6, /H6
* -- Comments (<!-- .... -->)
*
* Character tags:
*
* -- B, /B, STRONG, /STRONG
* -- I, /I, EM, /EM, VAR, /VAR, CITE, /CITE
* -- CODE, /CODE, SAMP, /SAMP, KBD, /KBD, TT, /TT
* -- U, /U
* -- STRIKE, /STRIKE
* -- CODE, /CODE
*
* The most obvious limitation is that neither tables
* nor frames are supported. Also forget about CSS
* and JavaScript, of course.
*
* All the ampersand (& something) sequences defined
* in HTML 3 are properly translated.
*
* Note: Those are translated to the ANSI (MS-Windows,
* OS/2 codepage 1004) character set. This has the
* following characteristics:
*
* -- Codes 0-127 are identical to ASCII and thus
* ISO 8559-1 ("Latin 1") also.
*
* -- Codes 160-255 are identical to ISO 8559-1 ("Latin 1").
*
* -- Codes 128-159 are NOT defined in ISO 8559-1, but
* Netscape treats those as ANSI as well, so we do too.
*
* As a result, consider the output to be in OS/2 codepage
* 1004. Either set your codepage to that (WinSetCp)
* or translate the output (WinCpTranslateString).
*
* &#xxx; tags (with xxx being a decimal) are considered
* ANSI codes as well. Even though HTML 4.0 allows Unicode
* characters > 255 to be inserted this way, we ignore
* those. Unicode chars from 0 to 255 are identical to
* ANSI, so for � to ÿ, we are HTML-compliant.
*
* All other tags are completely thrown out.
*
*@@added V0.9.3 (2000-05-06) [umoeller]
*@@changed V0.9.20 (2002-08-10) [umoeller]: changed prototype
*/
BOOL txvConvertFromHTML(PSZ *ppszText, // in/out: text (gets reallocated)
PSZ *ppszTitle, // out: if != NULL, receives malloc'd buffer with HTML title
PULONG pulProgress, // out: progress (ptr can be NULL)
PBOOL pfCancel) // in: cancel flag (ptr can be NULL)
{
BOOL brc = TRUE;
ULONG cbSource = strlen(*ppszText);
COPYTARGET ct = {0};
lstInit(&ct.llLists,
TRUE); // free items
ct.ppszTitle = ppszTitle; // V0.9.20 (2002-08-10) [umoeller]
// can be NULL
ct.pSource = *ppszText;
// skip leading spaces
ct.fSkipNextSpace = TRUE;
// step 2:
// actual tags formatting
while (TRUE)
{
CHAR c = *ct.pSource;
if (pfCancel)
if (*pfCancel)
{
brc = FALSE;
break;
}
if (!c)
// null terminator reached:
break;
// calculate progress
if (pulProgress)
*pulProgress = ((ct.pSource - *ppszText) // characters done
* 100
/ cbSource); // characters total
switch (c)
{
case '<':
HandleTag(&ct);
break;
case '&':
HandleEscape(&ct);
break;
case '\r':
// skip
if (!ct.fSkipNextSpace)
{
AppendChar(&ct,
' ');
// ct.fNeedsLinebreak = FALSE;
// but skip leading spaces which might follow
if (!ct.fPRE)
ct.fSkipNextSpace = TRUE;
}
ct.pSource++;
break;
case '\t':
{
if (ct.fPRE)
{
ULONG ul;
for (ul = 0;
ul < 8;
ul++)
AppendChar(&ct,
' ');
}
else
{
// not in PRE block:
if ( (!ct.fSkipNextSpace)
// && (!ct.fNeedsLinebreak)
)
// last was not space: copy
AppendChar(&ct,
' ');
ct.fSkipNextSpace = TRUE;
}
// skip the tab
ct.pSource++;
break; }
case '\n':
{
// newline char:
if (!ct.fPRE)
{
// if not in PRE mode, replace with space
if (!ct.fSkipNextSpace)
{
AppendChar(&ct,
' ');
// ct.fNeedsLinebreak = FALSE;
// but skip leading spaces which might follow
ct.fSkipNextSpace = TRUE;
}
}
else
// in PRE mode, preserve line breaks
AppendChar(&ct, '\n'); // ct.fNeedsLinebreak = TRUE;
ct.pSource++;
break; }
case '\xFF':
{
AppendChar(&ct,
' ');
ct.pSource++;
break; }
case ' ':
if (!ct.fPRE)
{
// is space, and not in PRE block:
if ( (!ct.fSkipNextSpace)
// && (!ct.fNeedsLinebreak)
)
// last was not space: copy
AppendChar(&ct,
' ');
ct.fSkipNextSpace = TRUE;
}
else
// in PRE, always add all spaces
AppendChar(&ct,
' ');
ct.pSource++;
break;
default:
// if we're not inserting escapes or anything,
// check if a linebreak is needed
AppendLinebreakCheck(&ct);
AppendChar(&ct,
*ct.pSource++);
ct.fSkipNextSpace = FALSE;
ct.fSkipNextLinebreak = FALSE;
} // end switch (*pSource);
} // end while (*pSource)
AppendChar(&ct,
'\n');
// append null-terminator
AppendChar(&ct,
0);
free(*ppszText);
*ppszText = ct.pszNew;
lstClear(&ct.llLists);
return brc;
}