From 5bc970f9a1ac9d2e79813a5f47afc896c49d0653 Mon Sep 17 00:00:00 2001 From: Ingo Schwarze Date: Tue, 18 Dec 2018 22:00:02 +0000 Subject: As a first step towards making roff_res() callable from mandoc_getarg(), move the function mandoc_getarg() from mandoc.c to roff.c. It was misplaced in mandoc.c in the first place; that file is intended for utilities needed both by parsers and by formatters, while reading macro arguments in copy mode is purely a task of the roff(7) parser. Needed as a preliminary for an upcoming bugfix. No code change. --- mandoc.c | 98 +-------------------------------------------------------------- roff.c | 99 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 99 insertions(+), 98 deletions(-) diff --git a/mandoc.c b/mandoc.c index 02bcf954..d1e7356b 100644 --- a/mandoc.c +++ b/mandoc.c @@ -1,4 +1,4 @@ -/* $Id: mandoc.c,v 1.112 2018/12/16 00:17:02 schwarze Exp $ */ +/* $Id: mandoc.c,v 1.113 2018/12/18 22:00:02 schwarze Exp $ */ /* * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons * Copyright (c) 2011-2015, 2017, 2018 Ingo Schwarze @@ -465,102 +465,6 @@ mandoc_escape(const char **end, const char **start, int *sz) return gly; } -/* - * Parse a quoted or unquoted roff-style request or macro argument. - * Return a pointer to the parsed argument, which is either the original - * pointer or advanced by one byte in case the argument is quoted. - * NUL-terminate the argument in place. - * Collapse pairs of quotes inside quoted arguments. - * Advance the argument pointer to the next argument, - * or to the NUL byte terminating the argument line. - */ -char * -mandoc_getarg(char **cpp, int ln, int *pos) -{ - char *start, *cp; - int quoted, pairs, white; - - /* Quoting can only start with a new word. */ - start = *cpp; - quoted = 0; - if ('"' == *start) { - quoted = 1; - start++; - } - - pairs = 0; - white = 0; - for (cp = start; '\0' != *cp; cp++) { - - /* - * Move the following text left - * after quoted quotes and after "\\" and "\t". - */ - if (pairs) - cp[-pairs] = cp[0]; - - if ('\\' == cp[0]) { - /* - * In copy mode, translate double to single - * backslashes and backslash-t to literal tabs. - */ - switch (cp[1]) { - case 'a': - case 't': - cp[0] = '\t'; - /* FALLTHROUGH */ - case '\\': - pairs++; - cp++; - break; - case ' ': - /* Skip escaped blanks. */ - if (0 == quoted) - cp++; - break; - default: - break; - } - } else if (0 == quoted) { - if (' ' == cp[0]) { - /* Unescaped blanks end unquoted args. */ - white = 1; - break; - } - } else if ('"' == cp[0]) { - if ('"' == cp[1]) { - /* Quoted quotes collapse. */ - pairs++; - cp++; - } else { - /* Unquoted quotes end quoted args. */ - quoted = 2; - break; - } - } - } - - /* Quoted argument without a closing quote. */ - if (1 == quoted) - mandoc_msg(MANDOCERR_ARG_QUOTE, ln, *pos, NULL); - - /* NUL-terminate this argument and move to the next one. */ - if (pairs) - cp[-pairs] = '\0'; - if ('\0' != *cp) { - *cp++ = '\0'; - while (' ' == *cp) - cp++; - } - *pos += (int)(cp - start) + (quoted ? 1 : 0); - *cpp = cp; - - if ('\0' == *cp && (white || ' ' == cp[-1])) - mandoc_msg(MANDOCERR_SPACE_EOL, ln, *pos, NULL); - - return start; -} - static int a2time(time_t *t, const char *fmt, const char *p) { diff --git a/roff.c b/roff.c index 96a5b33d..e25a74ef 100644 --- a/roff.c +++ b/roff.c @@ -1,4 +1,4 @@ -/* $Id: roff.c,v 1.352 2018/12/15 19:30:26 schwarze Exp $ */ +/* $Id: roff.c,v 1.353 2018/12/18 22:00:02 schwarze Exp $ */ /* * Copyright (c) 2008-2012, 2014 Kristaps Dzonsons * Copyright (c) 2010-2015, 2017, 2018 Ingo Schwarze @@ -1546,6 +1546,103 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos) return ROFF_CONT; } +/* + * Parse a quoted or unquoted roff-style request or macro argument. + * Return a pointer to the parsed argument, which is either the original + * pointer or advanced by one byte in case the argument is quoted. + * NUL-terminate the argument in place. + * Collapse pairs of quotes inside quoted arguments. + * Advance the argument pointer to the next argument, + * or to the NUL byte terminating the argument line. + */ +char * +mandoc_getarg(char **cpp, int ln, int *pos) +{ + char *start, *cp; + int quoted, pairs, white; + + /* Quoting can only start with a new word. */ + start = *cpp; + quoted = 0; + if ('"' == *start) { + quoted = 1; + start++; + } + + pairs = 0; + white = 0; + for (cp = start; '\0' != *cp; cp++) { + + /* + * Move the following text left + * after quoted quotes and after "\\" and "\t". + */ + if (pairs) + cp[-pairs] = cp[0]; + + if ('\\' == cp[0]) { + /* + * In copy mode, translate double to single + * backslashes and backslash-t to literal tabs. + */ + switch (cp[1]) { + case 'a': + case 't': + cp[0] = '\t'; + /* FALLTHROUGH */ + case '\\': + pairs++; + cp++; + break; + case ' ': + /* Skip escaped blanks. */ + if (0 == quoted) + cp++; + break; + default: + break; + } + } else if (0 == quoted) { + if (' ' == cp[0]) { + /* Unescaped blanks end unquoted args. */ + white = 1; + break; + } + } else if ('"' == cp[0]) { + if ('"' == cp[1]) { + /* Quoted quotes collapse. */ + pairs++; + cp++; + } else { + /* Unquoted quotes end quoted args. */ + quoted = 2; + break; + } + } + } + + /* Quoted argument without a closing quote. */ + if (1 == quoted) + mandoc_msg(MANDOCERR_ARG_QUOTE, ln, *pos, NULL); + + /* NUL-terminate this argument and move to the next one. */ + if (pairs) + cp[-pairs] = '\0'; + if ('\0' != *cp) { + *cp++ = '\0'; + while (' ' == *cp) + cp++; + } + *pos += (int)(cp - start) + (quoted ? 1 : 0); + *cpp = cp; + + if ('\0' == *cp && (white || ' ' == cp[-1])) + mandoc_msg(MANDOCERR_SPACE_EOL, ln, *pos, NULL); + + return start; +} + + /* * Process text streams. */ -- cgit v1.2.3-56-ge451