/* pprint.c -- pretty print parse tree (c) 1998-2007 (W3C) MIT, ERCIM, Keio University See tidy.h for the copyright notice. */ #include #include #include #include "pprint.h" #include "tidy-int.h" #include "parser.h" #include "entities.h" #include "tmbstr.h" #include "utf8.h" /* *** FOR DEBUG ONLY *** */ #if !defined(NDEBUG) && defined(_MSC_VER) /* #define DEBUG_PPRINT */ /* #define DEBUG_INDENT */ #ifdef DEBUG_PPRINT extern void dbg_show_node( TidyDocImpl* doc, Node *node, int caller, int indent ); #endif #ifdef DEBUG_INDENT #include "sprtf.h" #endif #endif /* Block-level and unknown elements are printed on new lines and their contents indented 2 spaces Inline elements are printed inline. Inline content is wrapped on spaces (except in attribute values or preformatted text, after start tags and before end tags */ static void PPrintAsp( TidyDocImpl* doc, uint indent, Node* node ); static void PPrintJste( TidyDocImpl* doc, uint indent, Node* node ); static void PPrintPhp( TidyDocImpl* doc, uint indent, Node* node ); static int TextEndsWithNewline( Lexer *lexer, Node *node, uint mode ); static int TextStartsWithWhitespace( Lexer *lexer, Node *node, uint start, uint mode ); static Bool InsideHead( TidyDocImpl* doc, Node *node ); static Bool ShouldIndent( TidyDocImpl* doc, Node *node ); /*\ * Issue #228 20150715 - macros to access --vertical-space tri state configuration parameter \*/ #define TidyClassicVS ((cfgAutoBool( doc, TidyVertSpace ) == TidyYesState) ? yes : no) #define TidyAddVS ((cfgAutoBool( doc, TidyVertSpace ) == TidyAutoState) ? no : yes ) /*\ * 20150515 - support using tabs instead of spaces - Issue #108 * GH: https://github.com/htacg/tidy-html5/issues/108 - Keep indent with tabs #108 * SF: https://sourceforge.net/p/tidy/feature-requests/3/ - #3 tabs in place of spaces \*/ static uint indent_char = ' '; void TY_(PPrintTabs)(void) { indent_char = '\t'; } void TY_(PPrintSpaces)(void) { indent_char = ' '; } #if SUPPORT_ASIAN_ENCODINGS /* #431953 - start RJ Wraplen adjusted for smooth international ride */ typedef enum { UC00, /* None */ UCPC, /* Punctuation, Connector */ UCPD, /* Punctuation, Dash */ UCPE, /* Punctuation, Close */ UCPS, /* Punctuation, Open */ UCPI, /* Punctuation, Initial quote */ UCPF, /* Punctuation, Final quote */ UCPO, /* Punctuation, Other */ UCZS, /* Separator, Space */ UCZL, /* Separator, Line */ UCZP /* Separator, Paragraph */ } UnicodeCategory; /* From the original code, the following characters are removed: U+2011 (non-breaking hyphen) U+202F (narrow non-break space) U+2044 (fraction slash) U+200B (zero width space) ...... (bidi formatting control characters) U+2011 and U+202F are non-breaking, U+2044 is a Sm character, U+200B is a non-visible space, wrapping after it would make this space visible, bidi should be done using HTML features and the characters are neither Px or Zx. The following Unicode 3.0 punctuation characters are added: U+2048 (question exclamation mark) U+2049 (exclamation question mark) U+204A (tironian sign et) U+204B (reversed pilcrow sign) U+204C (black leftwards bullet) U+204D (black rightwards bullet) U+3030 (wavy dash) U+30FB (katakana middle dot) U+FE63 (small hyphen-minus) U+FE68 (small reverse solidus) U+FF3F (fullwidth low line) U+FF5B (fullwidth left curly bracket) U+FF5D (fullwidth right curly bracket) Other additional characters were not included in Unicode 3.0. The table is based on Unicode 4.0. It must include only those characters marking a wrapping point, "before" if the general category is UCPS or UCPI, otherwise "after". */ static struct _unicode4cat { unsigned long code; UnicodeCategory category; } const unicode4cat[] = { #if 0 { 0x037E, UCPO }, { 0x0387, UCPO }, { 0x055A, UCPO }, { 0x055B, UCPO }, { 0x055C, UCPO }, { 0x055D, UCPO }, { 0x055E, UCPO }, { 0x055F, UCPO }, { 0x0589, UCPO }, { 0x058A, UCPD }, { 0x05BE, UCPO }, { 0x05C0, UCPO }, { 0x05C3, UCPO }, { 0x05F3, UCPO }, { 0x05F4, UCPO }, { 0x060C, UCPO }, { 0x060D, UCPO }, { 0x061B, UCPO }, { 0x061F, UCPO }, { 0x066A, UCPO }, { 0x066B, UCPO }, { 0x066C, UCPO }, { 0x066D, UCPO }, { 0x06D4, UCPO }, { 0x0700, UCPO }, { 0x0701, UCPO }, { 0x0702, UCPO }, { 0x0703, UCPO }, { 0x0704, UCPO }, { 0x0705, UCPO }, { 0x0706, UCPO }, { 0x0707, UCPO }, { 0x0708, UCPO }, { 0x0709, UCPO }, { 0x070A, UCPO }, { 0x070B, UCPO }, { 0x070C, UCPO }, { 0x070D, UCPO }, { 0x0964, UCPO }, { 0x0965, UCPO }, { 0x0970, UCPO }, { 0x0DF4, UCPO }, { 0x0E4F, UCPO }, { 0x0E5A, UCPO }, { 0x0E5B, UCPO }, { 0x0F04, UCPO }, { 0x0F05, UCPO }, { 0x0F06, UCPO }, { 0x0F07, UCPO }, { 0x0F08, UCPO }, { 0x0F09, UCPO }, { 0x0F0A, UCPO }, { 0x0F0B, UCPO }, { 0x0F0D, UCPO }, { 0x0F0E, UCPO }, { 0x0F0F, UCPO }, { 0x0F10, UCPO }, { 0x0F11, UCPO }, { 0x0F12, UCPO }, { 0x0F3A, UCPS }, { 0x0F3B, UCPE }, { 0x0F3C, UCPS }, { 0x0F3D, UCPE }, { 0x0F85, UCPO }, { 0x104A, UCPO }, { 0x104B, UCPO }, { 0x104C, UCPO }, { 0x104D, UCPO }, { 0x104E, UCPO }, { 0x104F, UCPO }, { 0x10FB, UCPO }, { 0x1361, UCPO }, { 0x1362, UCPO }, { 0x1363, UCPO }, { 0x1364, UCPO }, { 0x1365, UCPO }, { 0x1366, UCPO }, { 0x1367, UCPO }, { 0x1368, UCPO }, { 0x166D, UCPO }, { 0x166E, UCPO }, { 0x1680, UCZS }, { 0x169B, UCPS }, { 0x169C, UCPE }, { 0x16EB, UCPO }, { 0x16EC, UCPO }, { 0x16ED, UCPO }, { 0x1735, UCPO }, { 0x1736, UCPO }, { 0x17D4, UCPO }, { 0x17D5, UCPO }, { 0x17D6, UCPO }, { 0x17D8, UCPO }, { 0x17D9, UCPO }, { 0x17DA, UCPO }, { 0x1800, UCPO }, { 0x1801, UCPO }, { 0x1802, UCPO }, { 0x1803, UCPO }, { 0x1804, UCPO }, { 0x1805, UCPO }, { 0x1806, UCPD }, { 0x1807, UCPO }, { 0x1808, UCPO }, { 0x1809, UCPO }, { 0x180A, UCPO }, { 0x180E, UCZS }, { 0x1944, UCPO }, { 0x1945, UCPO }, #endif { 0x2000, UCZS }, { 0x2001, UCZS }, { 0x2002, UCZS }, { 0x2003, UCZS }, { 0x2004, UCZS }, { 0x2005, UCZS }, { 0x2006, UCZS }, { 0x2008, UCZS }, { 0x2009, UCZS }, { 0x200A, UCZS }, { 0x2010, UCPD }, { 0x2012, UCPD }, { 0x2013, UCPD }, { 0x2014, UCPD }, { 0x2015, UCPD }, { 0x2016, UCPO }, { 0x2017, UCPO }, { 0x2018, UCPI }, { 0x2019, UCPF }, { 0x201A, UCPS }, { 0x201B, UCPI }, { 0x201C, UCPI }, { 0x201D, UCPF }, { 0x201E, UCPS }, { 0x201F, UCPI }, { 0x2020, UCPO }, { 0x2021, UCPO }, { 0x2022, UCPO }, { 0x2023, UCPO }, { 0x2024, UCPO }, { 0x2025, UCPO }, { 0x2026, UCPO }, { 0x2027, UCPO }, { 0x2028, UCZL }, { 0x2029, UCZP }, { 0x2030, UCPO }, { 0x2031, UCPO }, { 0x2032, UCPO }, { 0x2033, UCPO }, { 0x2034, UCPO }, { 0x2035, UCPO }, { 0x2036, UCPO }, { 0x2037, UCPO }, { 0x2038, UCPO }, { 0x2039, UCPI }, { 0x203A, UCPF }, { 0x203B, UCPO }, { 0x203C, UCPO }, { 0x203D, UCPO }, { 0x203E, UCPO }, { 0x203F, UCPC }, { 0x2040, UCPC }, { 0x2041, UCPO }, { 0x2042, UCPO }, { 0x2043, UCPO }, { 0x2045, UCPS }, { 0x2046, UCPE }, { 0x2047, UCPO }, { 0x2048, UCPO }, { 0x2049, UCPO }, { 0x204A, UCPO }, { 0x204B, UCPO }, { 0x204C, UCPO }, { 0x204D, UCPO }, { 0x204E, UCPO }, { 0x204F, UCPO }, { 0x2050, UCPO }, { 0x2051, UCPO }, { 0x2053, UCPO }, { 0x2054, UCPC }, { 0x2057, UCPO }, { 0x205F, UCZS }, { 0x207D, UCPS }, { 0x207E, UCPE }, { 0x208D, UCPS }, { 0x208E, UCPE }, { 0x2329, UCPS }, { 0x232A, UCPE }, { 0x23B4, UCPS }, { 0x23B5, UCPE }, { 0x23B6, UCPO }, { 0x2768, UCPS }, { 0x2769, UCPE }, { 0x276A, UCPS }, { 0x276B, UCPE }, { 0x276C, UCPS }, { 0x276D, UCPE }, { 0x276E, UCPS }, { 0x276F, UCPE }, { 0x2770, UCPS }, { 0x2771, UCPE }, { 0x2772, UCPS }, { 0x2773, UCPE }, { 0x2774, UCPS }, { 0x2775, UCPE }, { 0x27E6, UCPS }, { 0x27E7, UCPE }, { 0x27E8, UCPS }, { 0x27E9, UCPE }, { 0x27EA, UCPS }, { 0x27EB, UCPE }, { 0x2983, UCPS }, { 0x2984, UCPE }, { 0x2985, UCPS }, { 0x2986, UCPE }, { 0x2987, UCPS }, { 0x2988, UCPE }, { 0x2989, UCPS }, { 0x298A, UCPE }, { 0x298B, UCPS }, { 0x298C, UCPE }, { 0x298D, UCPS }, { 0x298E, UCPE }, { 0x298F, UCPS }, { 0x2990, UCPE }, { 0x2991, UCPS }, { 0x2992, UCPE }, { 0x2993, UCPS }, { 0x2994, UCPE }, { 0x2995, UCPS }, { 0x2996, UCPE }, { 0x2997, UCPS }, { 0x2998, UCPE }, { 0x29D8, UCPS }, { 0x29D9, UCPE }, { 0x29DA, UCPS }, { 0x29DB, UCPE }, { 0x29FC, UCPS }, { 0x29FD, UCPE }, { 0x3001, UCPO }, { 0x3002, UCPO }, { 0x3003, UCPO }, { 0x3008, UCPS }, { 0x3009, UCPE }, { 0x300A, UCPS }, { 0x300B, UCPE }, { 0x300C, UCPS }, { 0x300D, UCPE }, { 0x300E, UCPS }, { 0x300F, UCPE }, { 0x3010, UCPS }, { 0x3011, UCPE }, { 0x3014, UCPS }, { 0x3015, UCPE }, { 0x3016, UCPS }, { 0x3017, UCPE }, { 0x3018, UCPS }, { 0x3019, UCPE }, { 0x301A, UCPS }, { 0x301B, UCPE }, { 0x301C, UCPD }, { 0x301D, UCPS }, { 0x301E, UCPE }, { 0x301F, UCPE }, { 0x3030, UCPD }, { 0x303D, UCPO }, { 0x30A0, UCPD }, { 0x30FB, UCPC }, { 0xFD3E, UCPS }, { 0xFD3F, UCPE }, { 0xFE30, UCPO }, { 0xFE31, UCPD }, { 0xFE32, UCPD }, { 0xFE33, UCPC }, { 0xFE34, UCPC }, { 0xFE35, UCPS }, { 0xFE36, UCPE }, { 0xFE37, UCPS }, { 0xFE38, UCPE }, { 0xFE39, UCPS }, { 0xFE3A, UCPE }, { 0xFE3B, UCPS }, { 0xFE3C, UCPE }, { 0xFE3D, UCPS }, { 0xFE3E, UCPE }, { 0xFE3F, UCPS }, { 0xFE40, UCPE }, { 0xFE41, UCPS }, { 0xFE42, UCPE }, { 0xFE43, UCPS }, { 0xFE44, UCPE }, { 0xFE45, UCPO }, { 0xFE46, UCPO }, { 0xFE47, UCPS }, { 0xFE48, UCPE }, { 0xFE49, UCPO }, { 0xFE4A, UCPO }, { 0xFE4B, UCPO }, { 0xFE4C, UCPO }, { 0xFE4D, UCPC }, { 0xFE4E, UCPC }, { 0xFE4F, UCPC }, { 0xFE50, UCPO }, { 0xFE51, UCPO }, { 0xFE52, UCPO }, { 0xFE54, UCPO }, { 0xFE55, UCPO }, { 0xFE56, UCPO }, { 0xFE57, UCPO }, { 0xFE58, UCPD }, { 0xFE59, UCPS }, { 0xFE5A, UCPE }, { 0xFE5B, UCPS }, { 0xFE5C, UCPE }, { 0xFE5D, UCPS }, { 0xFE5E, UCPE }, { 0xFE5F, UCPO }, { 0xFE60, UCPO }, { 0xFE61, UCPO }, { 0xFE63, UCPD }, { 0xFE68, UCPO }, { 0xFE6A, UCPO }, { 0xFE6B, UCPO }, { 0xFF01, UCPO }, { 0xFF02, UCPO }, { 0xFF03, UCPO }, { 0xFF05, UCPO }, { 0xFF06, UCPO }, { 0xFF07, UCPO }, { 0xFF08, UCPS }, { 0xFF09, UCPE }, { 0xFF0A, UCPO }, { 0xFF0C, UCPO }, { 0xFF0D, UCPD }, { 0xFF0E, UCPO }, { 0xFF0F, UCPO }, { 0xFF1A, UCPO }, { 0xFF1B, UCPO }, { 0xFF1F, UCPO }, { 0xFF20, UCPO }, { 0xFF3B, UCPS }, { 0xFF3C, UCPO }, { 0xFF3D, UCPE }, { 0xFF3F, UCPC }, { 0xFF5B, UCPS }, { 0xFF5D, UCPE }, { 0xFF5F, UCPS }, { 0xFF60, UCPE }, { 0xFF61, UCPO }, { 0xFF62, UCPS }, { 0xFF63, UCPE }, { 0xFF64, UCPO }, { 0xFF65, UCPC }, { 0x10100,UCPO }, { 0x10101,UCPO }, { 0x1039F,UCPO }, /* final entry */ { 0x0000, UC00 } }; typedef enum { NoWrapPoint, WrapBefore, WrapAfter } WrapPoint; /* If long lines of text have no white space as defined in HTML 4 (U+0009, U+000A, U+000D, U+000C, U+0020) other characters could be used to determine a wrap point. Since user agents would normalize the inserted newline character to a space character, this wrapping behaviour would insert visual whitespace into the document. Characters of the General Category Pi and Ps in the Unicode character database (opening punctuation and intial quote characters) mark a wrapping point before the character, other punctuation characters (Pc, Pd, Pe, Pf, and Po), breakable space characters (Zs), and paragraph and line separators (Zl, Zp) mark a wrap point after the character. Using this function Tidy can for example pretty print

....................“...quote...”...

as

....................\n“...quote...”...

or

....................“...quote...”\n...

if the next normal wrapping point would exceed the user chosen wrapping column. */ static WrapPoint CharacterWrapPoint(tchar c) { int i; for (i = 0; unicode4cat[i].code && unicode4cat[i].code <= c; ++i) if (unicode4cat[i].code == c) { /* wrapping before opening punctuation and initial quotes */ if (unicode4cat[i].category == UCPS || unicode4cat[i].category == UCPI) return WrapBefore; /* else wrapping after this character */ else return WrapAfter; } /* character has no effect on line wrapping */ return NoWrapPoint; } static WrapPoint Big5WrapPoint(tchar c) { if ((c & 0xFF00) == 0xA100) { /* opening brackets have odd codes: break before them */ if ( c > 0xA15C && c < 0xA1AD && (c & 1) == 1 ) return WrapBefore; return WrapAfter; } return NoWrapPoint; } #endif /* SUPPORT_ASIAN_ENCODINGS */ static void InitIndent( TidyIndent* ind ) { ind->spaces = -1; ind->attrValStart = -1; ind->attrStringStart = -1; } void TY_(InitPrintBuf)( TidyDocImpl* doc ) { TidyClearMemory( &doc->pprint, sizeof(TidyPrintImpl) ); InitIndent( &doc->pprint.indent[0] ); InitIndent( &doc->pprint.indent[1] ); doc->pprint.allocator = doc->allocator; doc->pprint.line = 0; } void TY_(FreePrintBuf)( TidyDocImpl* doc ) { TidyDocFree( doc, doc->pprint.linebuf ); TY_(InitPrintBuf)( doc ); } static void expand( TidyPrintImpl* pprint, uint len ) { uint* ip; uint buflen = pprint->lbufsize; if ( buflen == 0 ) buflen = 256; while ( len >= buflen ) buflen *= 2; ip = (uint*) TidyRealloc( pprint->allocator, pprint->linebuf, buflen*sizeof(uint) ); if ( ip ) { TidyClearMemory( ip+pprint->lbufsize, (buflen-pprint->lbufsize)*sizeof(uint) ); pprint->lbufsize = buflen; pprint->linebuf = ip; } } static uint GetSpaces( TidyPrintImpl* pprint ) { int spaces = pprint->indent[ 0 ].spaces; return ( spaces < 0 ? 0U : (uint) spaces ); } static int ClearInString( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + pprint->ixInd; return ind->attrStringStart = -1; } static int ToggleInString( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + pprint->ixInd; Bool inString = ( ind->attrStringStart >= 0 ); return ind->attrStringStart = ( inString ? -1 : (int) pprint->linelen ); } static Bool IsInString( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + 0; /* Always 1st */ return ( ind->attrStringStart >= 0 && ind->attrStringStart < (int) pprint->linelen ); } static Bool IsWrapInString( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + 0; /* Always 1st */ int wrap = (int) pprint->wraphere; return ( ind->attrStringStart == 0 || (ind->attrStringStart > 0 && ind->attrStringStart < wrap) ); } static Bool HasMixedContent (Node *element) { Node * node; if (!element) return no; for (node = element->content; node; node = node->next) if ( TY_(nodeIsText)(node) ) return yes; return no; } static void ClearInAttrVal( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + pprint->ixInd; ind->attrValStart = -1; } static int SetInAttrVal( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + pprint->ixInd; return ind->attrValStart = (int) pprint->linelen; } static Bool IsWrapInAttrVal( TidyPrintImpl* pprint ) { TidyIndent *ind = pprint->indent + 0; /* Always 1st */ int wrap = (int) pprint->wraphere; return ( ind->attrValStart == 0 || (ind->attrValStart > 0 && ind->attrValStart < wrap) ); } static Bool WantIndent( TidyDocImpl* doc ) { TidyPrintImpl* pprint = &doc->pprint; Bool wantIt = GetSpaces(pprint) > 0; if ( wantIt ) { Bool indentAttrs = cfgBool( doc, TidyIndentAttributes ); wantIt = ( ( !IsWrapInAttrVal(pprint) || indentAttrs ) && !IsWrapInString(pprint) ); } return wantIt; } static uint WrapOff( TidyDocImpl* doc ) { uint saveWrap = cfg( doc, TidyWrapLen ); TY_(SetOptionInt)( doc, TidyWrapLen, 0xFFFFFFFF ); /* very large number */ return saveWrap; } static void WrapOn( TidyDocImpl* doc, uint saveWrap ) { TY_(SetOptionInt)( doc, TidyWrapLen, saveWrap ); } static uint WrapOffCond( TidyDocImpl* doc, Bool onoff ) { if ( onoff ) return WrapOff( doc ); return cfg( doc, TidyWrapLen ); } static void AddC( TidyPrintImpl* pprint, uint c, uint string_index) { if ( string_index + 1 >= pprint->lbufsize ) expand( pprint, string_index + 1 ); pprint->linebuf[string_index] = c; } static uint AddChar( TidyPrintImpl* pprint, uint c ) { AddC( pprint, c, pprint->linelen ); return ++pprint->linelen; } static uint AddAsciiString( TidyPrintImpl* pprint, ctmbstr str, uint string_index ) { uint ix, len = TY_(tmbstrlen)( str ); if ( string_index + len >= pprint->lbufsize ) expand( pprint, string_index + len ); for ( ix=0; ixlinebuf[string_index + ix] = str[ ix ]; return string_index + len; } static uint AddString( TidyPrintImpl* pprint, ctmbstr str ) { return pprint->linelen = AddAsciiString( pprint, str, pprint->linelen ); } /* Saves current output point as the wrap point, ** but only if indentation would NOT overflow ** the current line. Otherwise keep previous wrap point. */ static Bool SetWrap( TidyDocImpl* doc, uint indent ) { TidyPrintImpl* pprint = &doc->pprint; Bool wrap = ( indent + pprint->linelen < cfg(doc, TidyWrapLen) ); if ( wrap ) { if ( pprint->indent[0].spaces < 0 ) pprint->indent[0].spaces = indent; pprint->wraphere = pprint->linelen; } else if ( pprint->ixInd == 0 ) { /* Save indent 1st time we pass the the wrap line */ pprint->indent[ 1 ].spaces = indent; pprint->ixInd = 1; } return wrap; } static void CarryOver( int* valTo, int* valFrom, uint wrapPoint ) { if ( *valFrom > (int) wrapPoint ) { *valTo = *valFrom - wrapPoint; *valFrom = -1; } } static Bool SetWrapAttr( TidyDocImpl* doc, uint indent, int attrStart, int strStart ) { TidyPrintImpl* pprint = &doc->pprint; TidyIndent *ind = pprint->indent + 0; Bool wrap = ( indent + pprint->linelen < cfg(doc, TidyWrapLen) ); if ( wrap ) { if ( ind[0].spaces < 0 ) ind[0].spaces = indent; pprint->wraphere = pprint->linelen; } else if ( pprint->ixInd == 0 ) { /* Save indent 1st time we pass the the wrap line */ pprint->indent[ 1 ].spaces = indent; pprint->ixInd = 1; /* Carry over string state */ CarryOver( &ind[1].attrStringStart, &ind[0].attrStringStart, pprint->wraphere ); CarryOver( &ind[1].attrValStart, &ind[0].attrValStart, pprint->wraphere ); } ind += doc->pprint.ixInd; ind->attrValStart = attrStart; ind->attrStringStart = strStart; return wrap; } /* Reset indent state after flushing a new line */ static void ResetLine( TidyPrintImpl* pprint ) { TidyIndent* ind = pprint->indent + 0; if ( pprint->ixInd > 0 ) { ind[0] = ind[1]; InitIndent( &ind[1] ); } if ( pprint->wraphere > 0 ) { int wrap = (int) pprint->wraphere; if ( ind[0].attrStringStart > wrap ) ind[0].attrStringStart -= wrap; if ( ind[0].attrValStart > wrap ) ind[0].attrValStart -= wrap; } else { if ( ind[0].attrStringStart > 0 ) ind[0].attrStringStart = 0; if ( ind[0].attrValStart > 0 ) ind[0].attrValStart = 0; } pprint->wraphere = pprint->ixInd = 0; } /* Shift text after wrap point to ** beginning of next line. */ static void ResetLineAfterWrap( TidyPrintImpl* pprint ) { if ( pprint->linelen > pprint->wraphere ) { uint *p = pprint->linebuf; uint *q = p + pprint->wraphere; uint *end = p + pprint->linelen; if ( ! IsWrapInAttrVal(pprint) ) { while ( q < end && *q == ' ' ) ++q, ++pprint->wraphere; } while ( q < end ) *p++ = *q++; pprint->linelen -= pprint->wraphere; } else { pprint->linelen = 0; } ResetLine( pprint ); } /*\ * Write the 'indent' char to output * Issue #335 - The GetSpaces() returns the number of spaces to be * used for the indent. This is fine if ouputting spaces. * However, if outputting 'tab' chars, then the number of tabs * output should euivalent to spaces divided by 'tab-size' \*/ static void WriteIndentChar(TidyDocImpl* doc ) { TidyPrintImpl* pprint = &doc->pprint; uint i; uint spaces = GetSpaces(pprint); uint tabsize = cfg(doc, TidyTabSize); if (spaces && (indent_char == '\t') && tabsize) { spaces /= tabsize; // set number of tabs to output if (spaces == 0) // with a minimum of one spaces = 1; } for (i = 0; i < spaces; i++) TY_(WriteChar)(indent_char, doc->docOut); /* 20150515 - Issue #108 */ } /* Goes ahead with writing current line up to ** previously saved wrap point. Shifts unwritten ** text in output buffer to beginning of next line. */ static void WrapLine( TidyDocImpl* doc ) { TidyPrintImpl* pprint = &doc->pprint; uint i; if ( pprint->wraphere == 0 ) return; if ( WantIndent(doc) ) WriteIndentChar(doc); for ( i = 0; i < pprint->wraphere; ++i ) TY_(WriteChar)( pprint->linebuf[i], doc->docOut ); if ( IsWrapInString(pprint) ) TY_(WriteChar)( '\\', doc->docOut ); TY_(WriteChar)( '\n', doc->docOut ); pprint->line++; ResetLineAfterWrap( pprint ); } /* Checks current output line length along with current indent. ** If combined they overflow output line length, go ahead ** and flush output up to the current wrap point. */ static Bool CheckWrapLine( TidyDocImpl* doc ) { TidyPrintImpl* pprint = &doc->pprint; if ( GetSpaces(pprint) + pprint->linelen >= cfg(doc, TidyWrapLen) ) { WrapLine( doc ); return yes; } return no; } static Bool CheckWrapIndent( TidyDocImpl* doc, uint indent ) { TidyPrintImpl* pprint = &doc->pprint; if ( GetSpaces(pprint) + pprint->linelen >= cfg(doc, TidyWrapLen) ) { WrapLine( doc ); if ( pprint->indent[ 0 ].spaces < 0 ) { #if !defined(NDEBUG) && defined(_MSC_VER) && defined(DEBUG_INDENT) SPRTF("%s Indent from %d to %d\n", __FUNCTION__, pprint->indent[ 0 ].spaces, indent ); #endif pprint->indent[ 0 ].spaces = indent; } return yes; } return no; } static void WrapAttrVal( TidyDocImpl* doc ) { TidyPrintImpl* pprint = &doc->pprint; uint i; /* assert( IsWrapInAttrVal(pprint) ); */ if ( WantIndent(doc) ) WriteIndentChar(doc); for ( i = 0; i < pprint->wraphere; ++i ) TY_(WriteChar)( pprint->linebuf[i], doc->docOut ); if ( IsWrapInString(pprint) ) TY_(WriteChar)( '\\', doc->docOut ); else TY_(WriteChar)( ' ', doc->docOut ); TY_(WriteChar)( '\n', doc->docOut ); pprint->line++; ResetLineAfterWrap( pprint ); } static void PFlushLineImpl( TidyDocImpl* doc ) { TidyPrintImpl* pprint = &doc->pprint; uint i; CheckWrapLine( doc ); if ( WantIndent(doc) ) WriteIndentChar(doc); for ( i = 0; i < pprint->linelen; ++i ) TY_(WriteChar)( pprint->linebuf[i], doc->docOut ); if ( IsInString(pprint) ) TY_(WriteChar)( '\\', doc->docOut ); ResetLine( pprint ); pprint->linelen = 0; } void TY_(PFlushLine)( TidyDocImpl* doc, uint indent ) { TidyPrintImpl* pprint = &doc->pprint; if ( pprint->linelen > 0 ) PFlushLineImpl( doc ); TY_(WriteChar)( '\n', doc->docOut ); pprint->line++; if (pprint->indent[ 0 ].spaces != (int)indent ) { #if !defined(NDEBUG) && defined(_MSC_VER) && defined(DEBUG_INDENT) SPRTF("%s Indent from %d to %d\n", __FUNCTION__, pprint->indent[ 0 ].spaces, indent ); #endif pprint->indent[ 0 ].spaces = indent; } } static void PCondFlushLine( TidyDocImpl* doc, uint indent ) { TidyPrintImpl* pprint = &doc->pprint; if ( pprint->linelen > 0 ) { PFlushLineImpl( doc ); TY_(WriteChar)( '\n', doc->docOut ); pprint->line++; } /* Issue #390 - Whether chars to flush or not, set new indent */ if ( pprint->indent[ 0 ].spaces != (int)indent ) { #if !defined(NDEBUG) && defined(_MSC_VER) && defined(DEBUG_INDENT) SPRTF("%s Indent from %d to %d\n", __FUNCTION__, pprint->indent[ 0 ].spaces, indent ); #endif pprint->indent[ 0 ].spaces = indent; } } /** * Two additional "smart" flush line functions which only * write a newline if `vertical-space no`. See issues #163 and #227. * These need to be used in the right place. In same cases `PFlushLine` * and `PCondFlushLine` should still be used. */ void TY_(PFlushLineSmart)( TidyDocImpl* doc, uint indent ) { TidyPrintImpl* pprint = &doc->pprint; if ( pprint->linelen > 0 ) PFlushLineImpl( doc ); /* Issue #228 - cfgBool( doc, TidyVertSpace ); */ if(TidyAddVS) { TY_(WriteChar)( '\n', doc->docOut ); pprint->line++; } if ( pprint->indent[ 0 ].spaces != (int)indent ) { #if !defined(NDEBUG) && defined(_MSC_VER) && defined(DEBUG_INDENT) SPRTF("%s Indent from %d to %d\n", __FUNCTION__, pprint->indent[ 0 ].spaces, indent ); #endif pprint->indent[ 0 ].spaces = indent; } } static void PCondFlushLineSmart( TidyDocImpl* doc, uint indent ) { TidyPrintImpl* pprint = &doc->pprint; if ( pprint->linelen > 0 ) { PFlushLineImpl( doc ); /* Issue #228 - cfgBool( doc, TidyVertSpace ); */ if(TidyAddVS) { TY_(WriteChar)( '\n', doc->docOut ); pprint->line++; } } /*\ * Issue #390 - Must still deal with fixing indent! * If TidyHideEndTags or TidyOmitOptionalTags, then * in certain circumstance no PrintEndTag will be done, * so linelen will be 0... \*/ if (pprint->indent[ 0 ].spaces != (int)indent) { #if !defined(NDEBUG) && defined(_MSC_VER) && defined(DEBUG_INDENT) SPRTF("%s Indent from %d to %d\n", __FUNCTION__, pprint->indent[ 0 ].spaces, indent ); #endif pprint->indent[ 0 ].spaces = indent; } } static void PPrintChar( TidyDocImpl* doc, uint c, uint mode ) { tmbchar entity[128]; ctmbstr p; TidyPrintImpl* pprint = &doc->pprint; uint outenc = cfg( doc, TidyOutCharEncoding ); Bool qmark = cfgBool( doc, TidyQuoteMarks ); if ( c == ' ' && !(mode & (PREFORMATTED | COMMENT | ATTRIBVALUE | CDATA))) { /* coerce a space character to a non-breaking space */ if (mode & NOWRAP) { ctmbstr ent = " "; /* by default XML doesn't define   */ if ( cfgBool(doc, TidyNumEntities) || cfgBool(doc, TidyXmlTags) ) ent = " "; AddString( pprint, ent ); return; } else pprint->wraphere = pprint->linelen; } /* comment characters are passed raw */ if ( mode & (COMMENT | CDATA) ) { AddChar( pprint, c ); return; } /* except in CDATA map < to < etc. */ if ( !(mode & CDATA) ) { if ( c == '<') { AddString( pprint, "<" ); return; } if ( c == '>') { AddString( pprint, ">" ); return; } /* naked '&' chars can be left alone or quoted as & The latter is required for XML where naked '&' are illegal. */ if ( c == '&' && cfgBool(doc, TidyQuoteAmpersand) && !cfgBool(doc, TidyPreserveEntities) && ( mode != OtherNamespace) ) /* #130 MathML attr and entity fix! */ { AddString( pprint, "&" ); return; } if ( c == '"' && qmark ) { AddString( pprint, """ ); return; } if ( c == '\'' && qmark ) { AddString( pprint, "'" ); return; } if ( c == 160 && outenc != RAW ) { if ( cfgBool(doc, TidyQuoteNbsp) ) { if ( cfgBool(doc, TidyNumEntities) || cfgBool(doc, TidyXmlTags) ) AddString( pprint, " " ); else AddString( pprint, " " ); } else AddChar( pprint, c ); return; } } #if SUPPORT_ASIAN_ENCODINGS /* #431953 - start RJ */ /* Handle encoding-specific issues */ switch ( outenc ) { case UTF8: #if SUPPORT_UTF16_ENCODINGS case UTF16: case UTF16LE: case UTF16BE: #endif if (!(mode & PREFORMATTED) && cfg(doc, TidyPunctWrap)) { WrapPoint wp = CharacterWrapPoint(c); if (wp == WrapBefore) pprint->wraphere = pprint->linelen; else if (wp == WrapAfter) pprint->wraphere = pprint->linelen + 1; } break; case BIG5: /* Allow linebreak at Chinese punctuation characters */ /* There are not many spaces in Chinese */ AddChar( pprint, c ); if (!(mode & PREFORMATTED) && cfg(doc, TidyPunctWrap)) { WrapPoint wp = Big5WrapPoint(c); if (wp == WrapBefore) pprint->wraphere = pprint->linelen; else if (wp == WrapAfter) pprint->wraphere = pprint->linelen + 1; } return; case SHIFTJIS: #ifndef NO_NATIVE_ISO2022_SUPPORT case ISO2022: /* ISO 2022 characters are passed raw */ #endif case RAW: AddChar( pprint, c ); return; } /* #431953 - end RJ */ #else /* SUPPORT_ASIAN_ENCODINGS */ /* otherwise ISO 2022 characters are passed raw */ if ( #ifndef NO_NATIVE_ISO2022_SUPPORT outenc == ISO2022 || #endif outenc == RAW ) { AddChar( pprint, c ); return; } #endif /* SUPPORT_ASIAN_ENCODINGS */ /* don't map latin-1 chars to entities */ if ( outenc == LATIN1 ) { if (c > 255) /* multi byte chars */ { uint vers = TY_(HTMLVersion)( doc ); if ( !cfgBool(doc, TidyNumEntities) && (p = TY_(EntityName)(c, vers)) ) TY_(tmbsnprintf)(entity, sizeof(entity), "&%s;", p); else TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c); AddString( pprint, entity ); return; } if (c > 126 && c < 160) { TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c); AddString( pprint, entity ); return; } AddChar( pprint, c ); return; } /* don't map UTF-8 chars to entities */ if ( outenc == UTF8 ) { AddChar( pprint, c ); return; } #if SUPPORT_UTF16_ENCODINGS /* don't map UTF-16 chars to entities */ if ( outenc == UTF16 || outenc == UTF16LE || outenc == UTF16BE ) { AddChar( pprint, c ); return; } #endif /* use numeric entities only for XML */ if ( cfgBool(doc, TidyXmlTags) ) { /* if ASCII use numeric entities for chars > 127 */ if ( c > 127 && outenc == ASCII ) { TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c); AddString( pprint, entity ); return; } /* otherwise output char raw */ AddChar( pprint, c ); return; } /* default treatment for ASCII */ if ( outenc == ASCII && (c > 126 || (c < ' ' && c != '\t')) ) { uint vers = TY_(HTMLVersion)( doc ); if (!cfgBool(doc, TidyNumEntities) && (p = TY_(EntityName)(c, vers)) ) TY_(tmbsnprintf)(entity, sizeof(entity), "&%s;", p); else TY_(tmbsnprintf)(entity, sizeof(entity), "&#%u;", c); AddString( pprint, entity ); return; } AddChar( pprint, c ); } static uint IncrWS( uint start, uint end, uint indent, int ixWS ) { if ( ixWS > 0 ) { uint st = start + MIN( (uint)ixWS, indent ); start = MIN( st, end ); } return start; } /* The line buffer is uint not char so we can hold Unicode values unencoded. The translation to UTF-8 is deferred to the TY_(WriteChar)() routine called to flush the line buffer. */ static void PPrintText( TidyDocImpl* doc, uint mode, uint indent, Node* node ) { uint start = node->start; uint end = node->end; uint ix, c = 0; int ixNL = TextEndsWithNewline( doc->lexer, node, mode ); int ixWS = TextStartsWithWhitespace( doc->lexer, node, start, mode ); if ( ixNL > 0 ) end -= ixNL; start = IncrWS( start, end, indent, ixWS ); for ( ix = start; ix < end; ++ix ) { CheckWrapIndent( doc, indent ); /* if ( CheckWrapIndent(doc, indent) ) { ixWS = TextStartsWithWhitespace( doc->lexer, node, ix ); ix = IncrWS( ix, end, indent, ixWS ); } */ c = (byte) doc->lexer->lexbuf[ix]; /* look for UTF-8 multibyte character */ if ( c > 0x7F ) ix += TY_(GetUTF8)( doc->lexer->lexbuf + ix, &c ); if ( c == '\n' ) { TY_(PFlushLine)( doc, indent ); ixWS = TextStartsWithWhitespace( doc->lexer, node, ix+1, mode ); ix = IncrWS( ix, end, indent, ixWS ); } else if (( c == '&' ) && (TY_(HTMLVersion)(doc) == HT50) && (((ix + 1) == end) || (((ix + 1) < end) && (isspace(doc->lexer->lexbuf[ix+1] & 0xff)))) ) { /*\ * Issue #207 - This is an unambiguous ampersand need not be 'quoted' in HTML5 * Issue #379 - Ensure only 0 to 255 passed to 'isspace' to avoid debug assert \*/ PPrintChar( doc, c, (mode | CDATA) ); } else { PPrintChar( doc, c, mode ); } } } #if 0 static void PPrintString( TidyDocImpl* doc, uint indent, ctmbstr str ) { while ( *str != '\0' ) AddChar( &doc->pprint, *str++ ); } #endif /* 0 */ static void PPrintAttrValue( TidyDocImpl* doc, uint indent, ctmbstr value, uint delim, Bool wrappable, Bool scriptAttr ) { TidyPrintImpl* pprint = &doc->pprint; Bool scriptlets = cfgBool(doc, TidyWrapScriptlets); uint mode = PREFORMATTED | ATTRIBVALUE; if ( wrappable ) mode = NORMAL | ATTRIBVALUE; /* look for ASP, Tango or PHP instructions for computed attribute value */ if ( value && value[0] == '<' ) { if ( value[1] == '%' || value[1] == '@'|| TY_(tmbstrncmp)(value, "wraphere > 0 && GetSpaces(pprint) + pprint->linelen >= wraplen ) WrapAttrVal( doc ); if ( c == delim ) { ctmbstr entity = (c == '"' ? """ : "'"); AddString( pprint, entity ); ++value; continue; } else if (c == '"') { if ( cfgBool(doc, TidyQuoteMarks) ) AddString( pprint, """ ); else AddChar( pprint, c ); if ( delim == '\'' && scriptAttr && scriptlets ) strStart = ToggleInString( pprint ); ++value; continue; } else if ( c == '\'' ) { if ( cfgBool(doc, TidyQuoteMarks) ) AddString( pprint, "'" ); else AddChar( pprint, c ); if ( delim == '"' && scriptAttr && scriptlets ) strStart = ToggleInString( pprint ); ++value; continue; } /* look for UTF-8 multibyte character */ if ( c > 0x7F ) value += TY_(GetUTF8)( value, &c ); ++value; if ( c == '\n' ) { /* No indent inside Javascript literals */ TY_(PFlushLine)( doc, (strStart < 0 && !cfgBool(doc, TidyLiteralAttribs) ? indent : 0) ); continue; } PPrintChar( doc, c, mode ); } ClearInAttrVal( pprint ); ClearInString( pprint ); } AddChar( pprint, delim ); } static uint AttrIndent( TidyDocImpl* doc, Node* node, AttVal* ARG_UNUSED(attr) ) { uint spaces = cfg( doc, TidyIndentSpaces ); uint xtra = 2; /* 1 for the '<', another for the ' ' */ if ( node->element == NULL ) return spaces; if ( !TY_(nodeHasCM)(node, CM_INLINE) || !ShouldIndent(doc, node->parent ? node->parent: node) ) return xtra + TY_(tmbstrlen)( node->element ); if ( NULL != (node = TY_(FindContainer)(node)) ) return xtra + TY_(tmbstrlen)( node->element ); return spaces; } static Bool AttrNoIndentFirst( /*TidyDocImpl* doc,*/ Node* node, AttVal* attr ) { return ( attr==node->attributes ); /*&& ( InsideHead(doc, node) || !TY_(nodeHasCM)(node, CM_INLINE) ) ); */ } static void PPrintAttribute( TidyDocImpl* doc, uint indent, Node *node, AttVal *attr ) { TidyPrintImpl* pprint = &doc->pprint; Bool xmlOut = cfgBool( doc, TidyXmlOut ); Bool xhtmlOut = cfgBool( doc, TidyXhtmlOut ); Bool wrapAttrs = cfgBool( doc, TidyWrapAttVals ); Bool ucAttrs = cfgBool( doc, TidyUpperCaseAttrs ); Bool indAttrs = cfgBool( doc, TidyIndentAttributes ); uint xtra = AttrIndent( doc, node, attr ); Bool first = AttrNoIndentFirst( /*doc,*/ node, attr ); tmbstr name = attr->attribute; Bool wrappable = no; tchar c; /* fix for odd attribute indentation bug triggered by long values */ if (!indAttrs) xtra = 0; if ( indAttrs ) { if ( TY_(nodeIsElement)(node) && !first ) { indent += xtra; PCondFlushLineSmart( doc, indent ); } else indAttrs = no; } CheckWrapIndent( doc, indent ); if ( !xmlOut && !xhtmlOut && attr->dict ) { if ( TY_(IsScript)(doc, name) ) wrappable = cfgBool( doc, TidyWrapScriptlets ); else if (!(attrIsCONTENT(attr) || attrIsVALUE(attr) || attrIsALT(attr) || attrIsTITLE(attr)) && wrapAttrs ) wrappable = yes; } if ( !first && !SetWrap(doc, indent) ) { TY_(PFlushLine)( doc, indent+xtra ); /* Put it on next line */ } else if ( pprint->linelen > 0 ) { AddChar( pprint, ' ' ); } /* Attribute name */ while (*name) { c = (unsigned char)*name; if (c > 0x7F) name += TY_(GetUTF8)(name, &c); else if (ucAttrs) c = TY_(ToUpper)(c); AddChar(pprint, c); ++name; } /* fix for bug 732038 */ #if 0 /* If not indenting attributes, bump up indent for ** value after putting out name. */ if ( !indAttrs ) indent += xtra; #endif CheckWrapIndent( doc, indent ); if ( attr->value == NULL ) { Bool isB = TY_(IsBoolAttribute)(attr); Bool scriptAttr = TY_(attrIsEvent)(attr); if ( xmlOut ) PPrintAttrValue( doc, indent, isB ? attr->attribute : NULLSTR, attr->delim, no, scriptAttr ); else if ( !isB && !TY_(IsNewNode)(node) ) PPrintAttrValue( doc, indent, "", attr->delim, yes, scriptAttr ); else SetWrap( doc, indent ); } else PPrintAttrValue( doc, indent, attr->value, attr->delim, wrappable, no ); } static void PPrintAttrs( TidyDocImpl* doc, uint indent, Node *node ) { TidyPrintImpl* pprint = &doc->pprint; AttVal* av; /* add xml:space attribute to pre and other elements */ if ( cfgBool(doc, TidyXmlOut) && cfgBool(doc, TidyXmlSpace) && !TY_(GetAttrByName)(node, "xml:space") && TY_(XMLPreserveWhiteSpace)(doc, node) ) { TY_(AddAttribute)( doc, node, "xml:space", "preserve" ); } for ( av = node->attributes; av; av = av->next ) { if ( av->attribute != NULL ) { PPrintAttribute( doc, indent, node, av ); } else if ( av->asp != NULL ) { AddChar( pprint, ' ' ); PPrintAsp( doc, indent, av->asp ); } else if ( av->php != NULL ) { AddChar( pprint, ' ' ); PPrintPhp( doc, indent, av->php ); } } } Bool TY_(TextNodeEndWithSpace)( Lexer *lexer, Node *node ) { if (TY_(nodeIsText)(node) && node->end > node->start) { uint i, c = '\0'; /* initialised to avoid warnings */ for (i = node->start; i < node->end; ++i) { c = (byte) lexer->lexbuf[i]; if ( c > 0x7F ) i += TY_(GetUTF8)( lexer->lexbuf + i, &c ); } if ( c == ' ' || c == '\n' ) return yes; } return no; } /* Line can be wrapped immediately after inline start tag provided if follows a text node ending in a space, or it follows a
, or its parent is an inline element that that rule applies to. This behaviour was reverse engineered from Netscape 3.0. Line wrapping can occur if an element is not empty and before a block level. For instance:

xy

will display properly. Whereas

x

won't. */ static Bool AfterSpaceImp(Lexer *lexer, Node *node, Bool isEmpty) { Node *prev; if ( !TY_(nodeCMIsInline)(node) ) return yes; prev = node->prev; if (prev) { if (TY_(nodeIsText)(prev)) return TY_(TextNodeEndWithSpace)( lexer, prev ); else if (nodeIsBR(prev)) return yes; return no; } if ( isEmpty && !TY_(nodeCMIsInline)(node->parent) ) return no; return AfterSpaceImp(lexer, node->parent, isEmpty); } static Bool AfterSpace(Lexer *lexer, Node *node) { return AfterSpaceImp(lexer, node, TY_(nodeCMIsEmpty)(node)); } static void PPrintEndTag( TidyDocImpl* doc, uint ARG_UNUSED(mode), uint ARG_UNUSED(indent), Node *node ); /*\ * See Issue #162 - void elements also get a closing tag, like img, br, ... * * from : http://www.w3.org/TR/html-markup/syntax.html#syntax-elements * A complete list of the void elements in HTML: * area, base, br, col, command, embed, hr, img, input, keygen, link, meta, param, source, track, wbr * * This could be sped up by NOT using the macro nodeIsXXXX, since this repeatedly checks the node, * and then the node->tag, which here are checked at the beginning... * * Some have already been done... at least where no macro yet exists. * * And maybe a switch(id) case would be faster. \*/ static Bool TY_(isVoidElement)( Node *node ) { TidyTagId id; if ( !node ) return no; if ( !node->tag ) return no; id = node->tag->id; if (nodeIsAREA(node)) return yes; if (nodeIsBASE(node)) return yes; if (nodeIsBR(node)) return yes; if (nodeIsCOL(node)) return yes; /* if (nodeIsCOMMAND(node)) */ if (id == TidyTag_COMMAND) return yes; if (nodeIsEMBED(node)) return yes; if (nodeIsHR(node)) return yes; if (nodeIsIMG(node)) return yes; if (nodeIsINPUT(node)) return yes; /* if (nodeIsKEYGEN(node)) */ if (id == TidyTag_KEYGEN ) return yes; if (nodeIsLINK(node)) return yes; if (nodeIsMETA(node)) return yes; if (nodeIsPARAM(node)) return yes; /* if (nodeIsSOURCE(node)) */ if (id == TidyTag_SOURCE ) return yes; /* if (nodeIsTRACK(node)) */ if (id == TidyTag_TRACK ) return yes; if (nodeIsWBR(node)) return yes; return no; } static void PPrintTag( TidyDocImpl* doc, uint mode, uint indent, Node *node ) { TidyPrintImpl* pprint = &doc->pprint; Bool uc = cfgBool( doc, TidyUpperCaseTags ); Bool xhtmlOut = cfgBool( doc, TidyXhtmlOut ); Bool xmlOut = cfgBool( doc, TidyXmlOut ); tchar c; tmbstr s = node->element; AddChar( pprint, '<' ); if ( node->type == EndTag ) AddChar( pprint, '/' ); if (s) { while (*s) { c = (unsigned char)*s; if (c > 0x7F) s += TY_(GetUTF8)(s, &c); else if (uc) c = TY_(ToUpper)(c); AddChar(pprint, c); ++s; } } PPrintAttrs( doc, indent, node ); if ( (xmlOut || xhtmlOut) && (node->type == StartEndTag || TY_(nodeCMIsEmpty)(node)) ) { AddChar( pprint, ' ' ); /* Space is NS compatibility hack
*/ AddChar( pprint, '/' ); /* Required end tag marker */ } AddChar( pprint, '>' ); /*\ * Appears this was added for Issue #111, #112, #113, but will now add an end tag * for elements like which do NOT have an EndTag, even in html5 * See Issue #162 - void elements also get a closing tag, like img, br, ... * A complete list of the void elements in HTML: * area, base, br, col, command, embed, hr, img, input, keygen, link, meta, param, source, track, wbr \*/ if ((node->type == StartEndTag && TY_(HTMLVersion)(doc) == HT50) && !TY_(isVoidElement)(node) ) { PPrintEndTag( doc, mode, indent, node ); } if ( (node->type != StartEndTag || xhtmlOut || (node->type == StartEndTag && TY_(HTMLVersion)(doc) == HT50)) && !(mode & PREFORMATTED) ) { uint wraplen = cfg( doc, TidyWrapLen ); CheckWrapIndent( doc, indent ); if ( indent + pprint->linelen < wraplen ) { /* wrap after start tag if is
or if it's not inline. Technically, it would be safe to call only AfterSpace. However, it would disrupt the existing algorithm. So let's leave as is. Note that AfterSpace returns true for non inline elements but can still be false for some
. So it has to stay as well. */ if (!(mode & NOWRAP) && (!TY_(nodeCMIsInline)(node) || nodeIsBR(node)) && AfterSpace(doc->lexer, node)) { pprint->wraphere = pprint->linelen; } } /* flush the current buffer only if it is known to be safe, i.e. it will not introduce some spurious white spaces. See bug #996484 */ else if ( mode & NOWRAP || nodeIsBR(node) || AfterSpace(doc->lexer, node)) PCondFlushLineSmart( doc, indent ); } } static void PPrintEndTag( TidyDocImpl* doc, uint ARG_UNUSED(mode), uint ARG_UNUSED(indent), Node *node ) { TidyPrintImpl* pprint = &doc->pprint; Bool uc = cfgBool( doc, TidyUpperCaseTags ); tmbstr s = node->element; tchar c; /* Netscape ignores SGML standard by not ignoring a line break before or etc. To avoid rendering this as an underlined space, I disable line wrapping before inline end tags by the #if 0 ... #endif */ #if 0 if ( !(mode & NOWRAP) ) SetWrap( doc, indent ); #endif AddString( pprint, " 0x7F) s += TY_(GetUTF8)(s, &c); else if (uc) c = TY_(ToUpper)(c); AddChar(pprint, c); ++s; } } AddChar( pprint, '>' ); } static void PPrintComment( TidyDocImpl* doc, uint indent, Node* node ) { TidyPrintImpl* pprint = &doc->pprint; SetWrap( doc, indent ); AddString( pprint, "