X-Git-Url: https://git.cameronkatri.com/mandoc.git/blobdiff_plain/5eec60b6e3415a34bd5237de153e27e9800dba8e..6cf3130222c8a052812f22a637ff37432d52c42f:/cgi.c
diff --git a/cgi.c b/cgi.c
index e4a31ada..91310ce4 100644
--- a/cgi.c
+++ b/cgi.c
@@ -1,19 +1,21 @@
-/* $Id: cgi.c,v 1.95 2014/08/21 16:05:21 schwarze Exp $ */
+/* $Id: cgi.c,v 1.175 2021/08/19 15:23:36 schwarze Exp $ */
/*
+ * Copyright (c) 2014-2019, 2021 Ingo Schwarze \n"
+ resp_begin_html(200, NULL, NULL);
+ resp_searchform(req, FOCUS_QUERY);
+ printf(" \n"
"This web interface is documented in the\n"
- "man.cgi\n"
+ "man.cgi(8)\n"
"manual, and the\n"
- "apropos\n"
+ "apropos(1)\n"
"manual explains the query syntax.\n"
- "
"); - puts(msg); - puts("
"); + resp_begin_html(code, http_msg, NULL); + resp_searchform(req, FOCUS_QUERY); + puts(""); + puts(user_msg); + puts("
"); resp_end_html(); } @@ -572,103 +585,101 @@ static void pg_error_badrequest(const char *msg) { - resp_begin_html(400, "Bad Request"); - puts("\n"); + resp_begin_html(400, "Bad Request", NULL); + puts("
\n"); puts(msg); printf("Try again from the\n" - "main page.\n" - "
", scriptname); + "main page.\n" + "", scriptname); resp_end_html(); } static void pg_error_internal(void) { - resp_begin_html(500, "Internal Server Error"); - puts("Internal Server Error
"); + resp_begin_html(500, "Internal Server Error", NULL); + puts("Internal Server Error
"); resp_end_html(); } +static void +pg_redirect(const struct req *req, const char *name) +{ + printf("Status: 303 See Other\r\n" + "Location: /"); + if (*scriptname != '\0') + printf("%s/", scriptname); + if (strcmp(req->q.manpath, req->p[0])) + printf("%s/", req->q.manpath); + if (req->q.arch != NULL) + printf("%s/", req->q.arch); + http_encode(name); + if (req->q.sec != NULL) { + putchar('.'); + http_encode(req->q.sec); + } + printf("\r\nContent-Type: text/html; charset=utf-8\r\n\r\n"); +} + static void pg_searchres(const struct req *req, struct manpage *r, size_t sz) { char *arch, *archend; - size_t i, iuse, isec; + const char *sec; + size_t i, iuse; int archprio, archpriouse; int prio, priouse; - char sec; for (i = 0; i < sz; i++) { if (validate_filename(r[i].file)) continue; - fprintf(stderr, "invalid filename %s in %s database\n", + warnx("invalid filename %s in %s database", r[i].file, req->q.manpath); pg_error_internal(); return; } - if (1 == sz) { + if (req->isquery && sz == 1) { /* * If we have just one result, then jump there now * without any delay. */ - printf("Status: 303 See Other\r\n"); - printf("Location: http://%s%s/%s/%s?", - HTTP_HOST, scriptname, req->q.manpath, r[0].file); - http_printquery(req, "&"); - printf("\r\n" - "Content-Type: text/html; charset=utf-8\r\n" - "\r\n"); + printf("Status: 303 See Other\r\n" + "Location: /"); + if (*scriptname != '\0') + printf("%s/", scriptname); + if (strcmp(req->q.manpath, req->p[0])) + printf("%s/", req->q.manpath); + printf("%s\r\n" + "Content-Type: text/html; charset=utf-8\r\n\r\n", + r[0].file); return; } - resp_begin_html(200, NULL); - resp_searchform(req); - puts("\n" - "q.manpath, r[i].file); - http_printquery(req, "&"); - printf("\">"); - html_print(r[i].names); - printf("\n" - " | \n" - ""); - html_print(r[i].output); - puts(" | \n" - "
" + "q.manpath, req->p[0])) + printf("%s/", req->q.manpath); + printf("%s\">", r[i].file); + html_print(r[i].names); + printf(" | \n" + ""); + html_print(r[i].output); + puts(" | \n" + "
You specified an invalid manual file.
"); + if ((f = fopen(file, "r")) == NULL) { + puts("You specified an invalid manual file.
"); return; } - puts(""); + puts("\n" + "\n" + ""); + puts(""); + + p = NULL; + sz = 0; - while (NULL != (p = fgetln(f, &len))) { + while ((len = getline(&p, &sz, f)) != -1) { bold = italic = 0; - for (i = 0; i < (int)len - 1; i++) { - /* + for (i = 0; i < len - 1; i++) { + /* * This means that the catpage is out of state. * Ignore it and keep going (although the * catpage is bogus). @@ -729,22 +774,22 @@ catman(const struct req *req, const char *file) if ('\b' != p[i + 1]) { if (italic) - printf(""); + printf(""); if (bold) - printf(""); + printf(""); italic = bold = 0; html_putchar(p[i]); continue; - } else if (i + 2 >= (int)len) + } else if (i + 2 >= len) continue; /* Italic mode. */ if ('_' == p[i]) { if (bold) - printf(""); + printf(""); if ( ! italic) - printf(""); + printf(""); bold = 0; italic = 1; i += 2; @@ -752,7 +797,7 @@ catman(const struct req *req, const char *file) continue; } - /* + /* * Handle funny behaviour troff-isms. * These grok'd from the original man2html.c. */ @@ -766,9 +811,9 @@ catman(const struct req *req, const char *file) ('*' == p[i] && '|' == p[i + 2]) || ('|' == p[i] && '*' == p[i + 2])) { if (italic) - printf(""); + printf(""); if (bold) - printf(""); + printf(""); italic = bold = 0; putchar('*'); i += 2; @@ -780,9 +825,9 @@ catman(const struct req *req, const char *file) ('+' == p[i] && '|' == p[i + 1]) || ('|' == p[i] && '+' == p[i + 1])) { if (italic) - printf(""); + printf(""); if (bold) - printf(""); + printf(""); italic = bold = 0; putchar('+'); i += 2; @@ -790,96 +835,81 @@ catman(const struct req *req, const char *file) } /* Bold mode. */ - + if (italic) - printf(""); + printf(""); if ( ! bold) - printf(""); + printf(""); bold = 1; italic = 0; i += 2; html_putchar(p[i]); } - /* + /* * Clean up the last character. - * We can get to a newline; don't print that. + * We can get to a newline; don't print that. */ if (italic) - printf(""); + printf(""); if (bold) - printf(""); + printf(""); - if (i == (int)len - 1 && '\n' != p[i]) + if (i == len - 1 && p[i] != '\n') html_putchar(p[i]); putchar('\n'); } + free(p); - puts("\n" - "
You specified an invalid manual file.
"); + puts("You specified an invalid manual file.
"); return; } - mp = mparse_alloc(MPARSE_SO, MANDOCLEVEL_FATAL, NULL, - req->q.manpath); - rc = mparse_readfd(mp, fd, file); + mchars_alloc(); + mp = mparse_alloc(MPARSE_SO | MPARSE_UTF8 | MPARSE_LATIN1 | + MPARSE_VALIDATE, MANDOC_OS_OTHER, req->q.manpath); + mparse_readfd(mp, fd, file); close(fd); + meta = mparse_result(mp); - if (rc >= MANDOCLEVEL_FATAL) { - fprintf(stderr, "fatal mandoc error: %s/%s\n", - req->q.manpath, file); - pg_error_internal(); - return; - } - + memset(&conf, 0, sizeof(conf)); + conf.fragment = 1; + conf.style = mandoc_strdup(CSS_DIR "/mandoc.css"); usepath = strcmp(req->q.manpath, req->p[0]); - mandoc_asprintf(&opts, - "fragment,man=%s?query=%%N&sec=%%S%s%s%s%s", - scriptname, - req->q.arch ? "&arch=" : "", - req->q.arch ? req->q.arch : "", - usepath ? "&manpath=" : "", - usepath ? req->q.manpath : ""); - - mparse_result(mp, &mdoc, &man, NULL); - if (NULL == man && NULL == mdoc) { - fprintf(stderr, "fatal mandoc error: %s/%s\n", - req->q.manpath, file); - pg_error_internal(); - mparse_free(mp); - return; - } - - vp = html_alloc(opts); + mandoc_asprintf(&conf.man, "/%s%s%s%s%%N.%%S", + scriptname, *scriptname == '\0' ? "" : "/", + usepath ? req->q.manpath : "", usepath ? "/" : ""); - if (NULL != mdoc) - html_mdoc(vp, mdoc); + vp = html_alloc(&conf); + if (meta->macroset == MACROSET_MDOC) + html_mdoc(vp, meta); else - html_man(vp, man); + html_man(vp, meta); html_free(vp); mparse_free(mp); - free(opts); + mchars_free(); + free(conf.man); + free(conf.style); } static void @@ -890,9 +920,9 @@ resp_show(const struct req *req, const char *file) file += 2; if ('c' == *file) - catman(req, file); + resp_catman(req, file); else - format(req, file); + resp_format(req, file); } static void @@ -905,7 +935,7 @@ pg_show(struct req *req, const char *fullpath) pg_error_badrequest( "You did not specify a page to show."); return; - } + } manpath = mandoc_strndup(fullpath, file - fullpath); file++; @@ -923,18 +953,12 @@ pg_show(struct req *req, const char *fullpath) */ if (chdir(manpath) == -1) { - fprintf(stderr, "chdir %s: %s\n", - manpath, strerror(errno)); + warn("chdir %s", manpath); pg_error_internal(); free(manpath); return; } - - if (strcmp(manpath, "mandoc")) { - free(req->q.manpath); - req->q.manpath = manpath; - } else - free(manpath); + free(manpath); if ( ! validate_filename(file)) { pg_error_badrequest( @@ -942,8 +966,8 @@ pg_show(struct req *req, const char *fullpath) return; } - resp_begin_html(200, NULL); - resp_searchform(req); + resp_begin_html(200, NULL, file); + resp_searchform(req, FOCUS_NONE); resp_show(req, file); resp_end_html(); } @@ -954,10 +978,10 @@ pg_search(const struct req *req) struct mansearch search; struct manpaths paths; struct manpage *res; - char **cp; - const char *ep, *start; + char **argv; + char *query, *rp, *wp; size_t ressz; - int i, sz; + int argc; /* * Begin by chdir()ing into the root of the manpath. @@ -965,9 +989,8 @@ pg_search(const struct req *req) * relative to the manpath root. */ - if (-1 == (chdir(req->q.manpath))) { - fprintf(stderr, "chdir %s: %s\n", - req->q.manpath, strerror(errno)); + if (chdir(req->q.manpath) == -1) { + warn("chdir %s", req->q.manpath); pg_error_internal(); return; } @@ -976,52 +999,57 @@ pg_search(const struct req *req) search.sec = req->q.sec; search.outkey = "Nd"; search.argmode = req->q.equal ? ARG_NAME : ARG_EXPR; + search.firstmatch = 1; paths.sz = 1; paths.paths = mandoc_malloc(sizeof(char *)); paths.paths[0] = mandoc_strdup("."); /* - * Poor man's tokenisation: just break apart by spaces. - * Yes, this is half-ass. But it works for now. + * Break apart at spaces with backslash-escaping. */ - ep = req->q.query; - while (ep && isspace((unsigned char)*ep)) - ep++; - - sz = 0; - cp = NULL; - while (ep && '\0' != *ep) { - cp = mandoc_reallocarray(cp, sz + 1, sizeof(char *)); - start = ep; - while ('\0' != *ep && ! isspace((unsigned char)*ep)) - ep++; - cp[sz] = mandoc_malloc((ep - start) + 1); - memcpy(cp[sz], start, ep - start); - cp[sz++][ep - start] = '\0'; - while (isspace((unsigned char)*ep)) - ep++; + argc = 0; + argv = NULL; + rp = query = mandoc_strdup(req->q.query); + for (;;) { + while (isspace((unsigned char)*rp)) + rp++; + if (*rp == '\0') + break; + argv = mandoc_reallocarray(argv, argc + 1, sizeof(char *)); + argv[argc++] = wp = rp; + for (;;) { + if (isspace((unsigned char)*rp)) { + *wp = '\0'; + rp++; + break; + } + if (rp[0] == '\\' && rp[1] != '\0') + rp++; + if (wp != rp) + *wp = *rp; + if (*rp == '\0') + break; + wp++; + rp++; + } } - if (0 == mansearch(&search, &paths, sz, cp, &res, &ressz)) - pg_noresult(req, "You entered an invalid query."); - else if (0 == ressz) - pg_noresult(req, "No results found."); + res = NULL; + ressz = 0; + if (req->isquery && req->q.equal && argc == 1) + pg_redirect(req, argv[0]); + else if (mansearch(&search, &paths, argc, argv, &res, &ressz) == 0) + pg_noresult(req, 400, "Bad Request", + "You entered an invalid query."); + else if (ressz == 0) + pg_noresult(req, 404, "Not Found", "No results found."); else pg_searchres(req, res, ressz); - for (i = 0; i < sz; i++) - free(cp[i]); - free(cp); - - for (i = 0; i < (int)ressz; i++) { - free(res[i].file); - free(res[i].names); - free(res[i].output); - } - free(res); - + free(query); + mansearch_free(res, ressz); free(paths.paths[0]); free(paths.paths); } @@ -1035,28 +1063,32 @@ main(void) const char *querystring; int i; - /* Poor man's ReDoS mitigation. */ +#if HAVE_PLEDGE + /* + * The "rpath" pledge could be revoked after mparse_readfd() + * if the file desciptor to "/footer.html" would be opened + * up front, but it's probably not worth the complication + * of the code it would cause: it would require scattering + * pledge() calls in multiple low-level resp_*() functions. + */ - itimer.it_value.tv_sec = 1; - itimer.it_value.tv_usec = 0; - itimer.it_interval.tv_sec = 1; - itimer.it_interval.tv_usec = 0; - if (setitimer(ITIMER_VIRTUAL, &itimer, NULL) == -1) { - fprintf(stderr, "setitimer: %s\n", strerror(errno)); + if (pledge("stdio rpath", NULL) == -1) { + warn("pledge"); pg_error_internal(); - return(EXIT_FAILURE); + return EXIT_FAILURE; } +#endif - /* Scan our run-time environment. */ - - if (NULL == (scriptname = getenv("SCRIPT_NAME"))) - scriptname = ""; + /* Poor man's ReDoS mitigation. */ - if ( ! validate_urifrag(scriptname)) { - fprintf(stderr, "unsafe SCRIPT_NAME \"%s\"\n", - scriptname); + itimer.it_value.tv_sec = 2; + itimer.it_value.tv_usec = 0; + itimer.it_interval.tv_sec = 2; + itimer.it_interval.tv_usec = 0; + if (setitimer(ITIMER_VIRTUAL, &itimer, NULL) == -1) { + warn("setitimer"); pg_error_internal(); - return(EXIT_FAILURE); + return EXIT_FAILURE; } /* @@ -1065,42 +1097,49 @@ main(void) * relative to the same position. */ - if (-1 == chdir(MAN_DIR)) { - fprintf(stderr, "MAN_DIR: %s: %s\n", - MAN_DIR, strerror(errno)); + if (chdir(MAN_DIR) == -1) { + warn("MAN_DIR: %s", MAN_DIR); pg_error_internal(); - return(EXIT_FAILURE); - } + return EXIT_FAILURE; + } memset(&req, 0, sizeof(struct req)); - pathgen(&req); + req.q.equal = 1; + parse_manpath_conf(&req); + + /* Parse the path info and the query string. */ + + if ((path = getenv("PATH_INFO")) == NULL) + path = ""; + else if (*path == '/') + path++; - /* Next parse out the query string. */ + if (*path != '\0') { + parse_path_info(&req, path); + if (req.q.manpath == NULL || req.q.sec == NULL || + *req.q.query == '\0' || access(path, F_OK) == -1) + path = ""; + } else if ((querystring = getenv("QUERY_STRING")) != NULL) + parse_query_string(&req, querystring); - if (NULL != (querystring = getenv("QUERY_STRING"))) - http_parse(&req, querystring); + /* Validate parsed data and add defaults. */ - if ( ! (NULL == req.q.manpath || - validate_manpath(&req, req.q.manpath))) { + if (req.q.manpath == NULL) + req.q.manpath = mandoc_strdup(req.p[0]); + else if ( ! validate_manpath(&req, req.q.manpath)) { pg_error_badrequest( "You specified an invalid manpath."); - return(EXIT_FAILURE); + return EXIT_FAILURE; } - if ( ! (NULL == req.q.arch || validate_urifrag(req.q.arch))) { + if (req.q.arch != NULL && validate_arch(req.q.arch) == 0) { pg_error_badrequest( "You specified an invalid architecture."); - return(EXIT_FAILURE); + return EXIT_FAILURE; } /* Dispatch to the three different pages. */ - path = getenv("PATH_INFO"); - if (NULL == path) - path = ""; - else if ('/' == *path) - path++; - if ('\0' != *path) pg_show(&req, path); else if (NULL != req.q.query) @@ -1115,49 +1154,125 @@ main(void) for (i = 0; i < (int)req.psz; i++) free(req.p[i]); free(req.p); - return(EXIT_SUCCESS); + return EXIT_SUCCESS; +} + +/* + * Translate PATH_INFO to a query. + */ +static void +parse_path_info(struct req *req, const char *path) +{ + const char *name, *sec, *end; + + req->isquery = 0; + req->q.equal = 1; + req->q.manpath = NULL; + req->q.arch = NULL; + + /* Mandatory manual page name. */ + if ((name = strrchr(path, '/')) == NULL) + name = path; + else + name++; + + /* Optional trailing section. */ + sec = strrchr(name, '.'); + if (sec != NULL && isdigit((unsigned char)*++sec)) { + req->q.query = mandoc_strndup(name, sec - name - 1); + req->q.sec = mandoc_strdup(sec); + } else { + req->q.query = mandoc_strdup(name); + req->q.sec = NULL; + } + + /* Handle the case of name[.section] only. */ + if (name == path) + return; + + /* Optional manpath. */ + end = strchr(path, '/'); + req->q.manpath = mandoc_strndup(path, end - path); + if (validate_manpath(req, req->q.manpath)) { + path = end + 1; + if (name == path) + return; + } else { + free(req->q.manpath); + req->q.manpath = NULL; + } + + /* Optional section. */ + if (strncmp(path, "man", 3) == 0 || strncmp(path, "cat", 3) == 0) { + path += 3; + end = strchr(path, '/'); + free(req->q.sec); + req->q.sec = mandoc_strndup(path, end - path); + path = end + 1; + if (name == path) + return; + } + + /* Optional architecture. */ + end = strchr(path, '/'); + if (end + 1 != name) { + pg_error_badrequest( + "You specified too many directory components."); + exit(EXIT_FAILURE); + } + req->q.arch = mandoc_strndup(path, end - path); + if (validate_arch(req->q.arch) == 0) { + pg_error_badrequest( + "You specified an invalid directory component."); + exit(EXIT_FAILURE); + } } /* * Scan for indexable paths. */ static void -pathgen(struct req *req) +parse_manpath_conf(struct req *req) { FILE *fp; char *dp; size_t dpsz; + ssize_t len; - if (NULL == (fp = fopen("manpath.conf", "r"))) { - fprintf(stderr, "%s/manpath.conf: %s\n", - MAN_DIR, strerror(errno)); + if ((fp = fopen("manpath.conf", "r")) == NULL) { + warn("%s/manpath.conf", MAN_DIR); pg_error_internal(); exit(EXIT_FAILURE); } - while (NULL != (dp = fgetln(fp, &dpsz))) { - if ('\n' == dp[dpsz - 1]) - dpsz--; + dp = NULL; + dpsz = 0; + + while ((len = getline(&dp, &dpsz, fp)) != -1) { + if (dp[len - 1] == '\n') + dp[--len] = '\0'; req->p = mandoc_realloc(req->p, (req->psz + 1) * sizeof(char *)); - dp = mandoc_strndup(dp, dpsz); if ( ! validate_urifrag(dp)) { - fprintf(stderr, "%s/manpath.conf contains " - "unsafe path \"%s\"\n", MAN_DIR, dp); + warnx("%s/manpath.conf contains " + "unsafe path \"%s\"", MAN_DIR, dp); pg_error_internal(); exit(EXIT_FAILURE); } - if (NULL != strchr(dp, '/')) { - fprintf(stderr, "%s/manpath.conf contains " - "path with slash \"%s\"\n", MAN_DIR, dp); + if (strchr(dp, '/') != NULL) { + warnx("%s/manpath.conf contains " + "path with slash \"%s\"", MAN_DIR, dp); pg_error_internal(); exit(EXIT_FAILURE); } req->p[req->psz++] = dp; + dp = NULL; + dpsz = 0; } + free(dp); - if ( req->p == NULL ) { - fprintf(stderr, "%s/manpath.conf is empty\n", MAN_DIR); + if (req->p == NULL) { + warnx("%s/manpath.conf is empty", MAN_DIR); pg_error_internal(); exit(EXIT_FAILURE); }