commit f0bc377e7d6b57bfcfc86fd61c2fa648cd225a33 (patch)
parent 3c5a5ae8f80a76f8ad41a82b1ff8dee2bf046d6a
Author: Alex Karle <alex@alexkarle.com>
Date: Thu, 30 Dec 2021 00:39:02 -0500
style: Convert whole source to adhere to style(9)
The OpenBSD style(9) guide seems as good a guide as any to adhere to,
since it's the OS I'm primarily developing on / targetting.
The non-curly else-if's after a multi-line if feel a bit weird/unsafe
(and I'm a tad surprised they wouldn't suggest always using parens to
combat easy mistakes.. I guess that's what compiler warnings are for?).
No real functional changes, although I basically get rid of parse()
since I realized handlec was basically doing all the heavy lifting...
Diffstat:
| M | nihdoc.c | | | 537 | +++++++++++++++++++++++++++++++++++++++++-------------------------------------- |
1 file changed, 277 insertions(+), 260 deletions(-)
diff --git a/nihdoc.c b/nihdoc.c
@@ -10,31 +10,28 @@
#include <unistd.h>
#include <stdbool.h>
-#define LNK_DES_MAX 2048
-
-
/* Global Constants and Enums */
char *FMT_STRS[] = {
- ['_'] = "em",
- ['*'] = "strong",
- ['`'] = "code",
+ ['_'] = "em",
+ ['*'] = "strong",
+ ['`'] = "code",
};
enum Block {
- NONE,
- HEADER,
- HEADER_PARSE,
- PARAGRAPH,
- CODE,
- LIST,
- LIST_PARSE,
+ NONE,
+ HEADER,
+ HEADER_PARSE,
+ PARAGRAPH,
+ CODE,
+ LIST,
+ LIST_PARSE,
};
enum Link {
- NOL,
- URL_PARSE,
- DESC_PARSE,
- OPT_URL,
+ NOL,
+ URL_PARSE,
+ DESC_PARSE,
+ OPT_URL,
};
@@ -44,7 +41,7 @@ enum Link in_link = NOL;
int hlvl = 0;
bool fmts[256] = {false}; /* indexed by _ * ` */
bool escape = false;
-char lnkdes[LNK_DES_MAX] = {0};
+char lnkdes[2048] = {0};
int lnkidx = 0;
int indent = 0;
int previndent = 0;
@@ -56,271 +53,291 @@ bool blockquote = false;
/* Helper functions */
-void putesc(int c) {
- switch (c) {
- case '<': printf("<"); break;
- case '>': printf(">"); break;
- case '&': printf("&"); break;
- default: putchar(c);
- }
+void
+putesc(int c)
+{
+ switch (c) {
+ case '<':
+ printf("<");
+ break;
+ case '>':
+ printf(">");
+ break;
+ case '&':
+ printf("&");
+ break;
+ default:
+ putchar(c);
+ }
}
-void newlist() {
- in = LIST;
- previndent = indent;
- printf("<%s>\n<li>\n", ol ? "ol" : "ul");
- listdepth++;
+void
+newlist(void)
+{
+ in = LIST;
+ previndent = indent;
+ printf("<%s>\n<li>\n", ol ? "ol" : "ul");
+ listdepth++;
}
-int endlist() {
- in = LIST;
- previndent = indent;
- printf("</li>\n</%s>\n", ol ? "ol" : "ul");
- return --listdepth;
+int
+endlist(void)
+{
+ in = LIST;
+ previndent = indent;
+ printf("</li>\n</%s>\n", ol ? "ol" : "ul");
+ return --listdepth;
}
-void maybe_startp() {
- /* All inline types should start the paragraph if no other major type present*/
- if (in == NONE) {
- in = PARAGRAPH;
- printf("<p>\n");
- }
+/*
+ * All inline types should start the paragraph if no other major type present.
+ * This function does just that, IFF we aren't already in a major type.
+ */
+void
+maybe_startp(void)
+{
+ if (in == NONE) {
+ in = PARAGRAPH;
+ printf("<p>\n");
+ }
}
-void handle_lf() {
- indent = 0;
- linestarted = false;
+void
+handle_lf(void)
+{
+ indent = 0;
+ linestarted = false;
- /* single line types (one lf to close) */
- if (in == HEADER) {
- in = NONE;
- printf("</h%d>\n", hlvl);
- }
+ /* single line types (one lf to close) */
+ if (in == HEADER) {
+ in = NONE;
+ printf("</h%d>\n", hlvl);
+ }
- /* multi-line types (two lf to close) */
- if (lastc == '\n' || (lastc == '>' && blockquote)) {
- switch (in) {
- case PARAGRAPH: printf("</p>\n"); break;
- case CODE: printf("</code></pre>\n"); break;
- case LIST:
- previndent = 0;
- while (endlist())
- ;
- break;
- default:
- break; /* no op */
- }
- in = NONE;
- if (blockquote && lastc == '\n') {
- printf("</blockquote>\n");
- blockquote = false;
- }
- }
+ /* multi-line types (two lf to close) */
+ if (lastc == '\n' || (lastc == '>' && blockquote)) {
+ switch (in) {
+ case PARAGRAPH:
+ printf("</p>\n");
+ break;
+ case CODE:
+ printf("</code></pre>\n");
+ break;
+ case LIST:
+ previndent = 0;
+ while (endlist())
+ ;
+ break;
+ default:
+ break; /* no op */
+ }
+ in = NONE;
+ if (blockquote && lastc == '\n') {
+ printf("</blockquote>\n");
+ blockquote = false;
+ }
+ }
}
-bool fmt_disabled(int c) {
- /* `` blocks disable all but the next `, likewise CODE makes all disabled */
- if (in == CODE || in_link == URL_PARSE) {
- return true;
- } else {
- return fmts['`'] && c != '`';
- }
+/* Returns whether inline styles should be allowed at this moment. */
+bool
+fmt_disabled(int c)
+{
+ /* `` blocks all but the next `, likewise CODE makes all disabled */
+ if (in == CODE || in_link == URL_PARSE)
+ return true;
+ else
+ return fmts['`'] && c != '`';
}
-void toggle_format(int c) {
- if (!fmt_disabled(c)) {
- maybe_startp();
- printf("<%s%s>", fmts[c] ? "/" : "", FMT_STRS[c]);
- fmts[c] ^= true;
- } else {
- putesc(c);
- }
+void
+toggle_format(int c)
+{
+ if (!fmt_disabled(c)) {
+ maybe_startp();
+ printf("<%s%s>", fmts[c] ? "/" : "", FMT_STRS[c]);
+ fmts[c] ^= true;
+ } else
+ putesc(c);
}
-void handlec(int c) {
- /* the main state machine, abstracted into a function to allow
- * playback support for buffered types (i.e. link descriptions) */
- /* Handle Escapes before any other bit of the main switch */
- if (escape) {
- maybe_startp();
- putesc(c);
- escape = false;
- return;
- }
+/*
+ * The main state machine [1], abstracted into a function to allow
+ * playback support for buffered types (i.e. link descriptions)
+ *
+ * [1]: home grown spaghetti code
+ */
+void
+handlec(int c)
+{
+ /*
+ * Store link descriptions as we go, skipping the regular loop since
+ * we'll play them back later via recursion on handlec
+ */
+ if (in_link == DESC_PARSE && c != ']') {
+ lnkdes[lnkidx++] = c;
+ return;
+ }
- /* Store the indentation */
- if (!linestarted && c == ' ') {
- indent++;
- return; /* don't print leading indents */
- }
+ /* Any character other than a '(' terminates a link at ']' */
+ if (in_link == OPT_URL && c != '(') {
+ in_link = NOL;
+ lnkdes[lnkidx] = '\0';
+ printf("<a href=\"%s\">%s</a>", lnkdes, lnkdes);
+ }
- switch (c) {
- case '\\':
- escape = true;
- break;
- case '#':
- if (in == NONE) {
- in = HEADER_PARSE;
- hlvl = 1;
- } else if (in == HEADER_PARSE) {
- hlvl++;
- } else {
- /* not a special # */
- putesc(c);
- }
- break;
- case ' ':
- if (in == HEADER_PARSE) {
- printf("<h%d>", hlvl);
- in = HEADER;
- } else if (in == LIST_PARSE) {
- if (!listdepth) {
- newlist();
- } else {
- if (previndent < indent) {
- newlist();
- } else if (previndent > indent) {
- endlist();
- printf("</li>\n<li>\n");
- } else {
- in = LIST;
- printf("</li>\n<li>\n");
- }
- }
- } else {
- putesc(c);
- }
- break;
- case '*':
- case '`':
- case '_':
- toggle_format(c);
- break;
- case '\t':
- if (in == NONE) {
- in = CODE;
- printf("<pre><code>");
- } else if (lastc == '\n' && c == '\t') {
- /* no op */
- } else {
- putesc(c);
- }
- break;
- case '>':
- if (in == NONE) {
- in = PARAGRAPH;
- /* only start a new blockquote block is not already in one */
- printf("%s", blockquote ? "<p>\n" : "<blockquote>\n<p>\n");
- blockquote = true;
- } else if (lastc == '\n' && c == '>') {
- /* no op */
- } else {
- putesc(c);
- }
- break;
- case '[':
- if (in_link == NOL && !fmt_disabled(c)) {
- maybe_startp();
- in_link = DESC_PARSE;
- lnkidx = 0;
- } else {
- putesc(c);
- }
- break;
- case ']':
- if (in_link == DESC_PARSE) {
- in_link = OPT_URL;
- } else {
- putesc(c);
- }
- break;
- case '(':
- if (in_link == OPT_URL) {
- /* don't print it, start parsing */
- in_link = URL_PARSE;
- printf("<a href=\"");
- } else {
- putesc(c);
- }
- break;
- case ')':
- if (in_link == URL_PARSE) {
- in_link = NOL;
- printf("\">");
- for (int i = 0; i < lnkidx; i++) {
- handlec(lnkdes[i]);
- }
- printf("</a>");
- } else {
- putesc(c);
- }
- break;
- case '1':
- case '2':
- case '3':
- case '4':
- case '5':
- case '6':
- case '7':
- case '8':
- case '9':
- case '.':
- case '-':
- if (in == NONE || !linestarted) {
- ol = c != '-';
- in = LIST_PARSE;
- } else if (in != LIST_PARSE) {
- putesc(c);
- }
- break;
- case '\n':
- handle_lf();
- if (in != NONE) {
- putesc(c);
- }
- break;
- default:
- maybe_startp();
- putesc(c);
- break;
- }
- lastc = c;
- linestarted = c != '\n';
-}
+ /* Handle Escapes before any other bit of the main switch */
+ if (escape) {
+ maybe_startp();
+ putesc(c);
+ escape = false;
+ return;
+ }
-int parse() {
- /* Mini state machine (home grown spaghetti code) */
- int c;
- while ((c = getchar()) != EOF) {
- /* Store links as we go, before the main loop */
- if (in_link == DESC_PARSE && c != ']') {
- lnkdes[lnkidx++] = c;
- /* skip the regular loop -- it'll be played back later */
- continue;
- }
+ /* Store the indentation and return without printing */
+ if (!linestarted && c == ' ') {
+ indent++;
+ return;
+ }
- /* Any character other than a '(' terminates a link at ']' */
- if (in_link == OPT_URL && c != '(') {
- in_link = NOL;
- lnkdes[lnkidx] = '\0';
- printf("<a href=\"%s\">%s</a>", lnkdes, lnkdes);
- }
-
- handlec(c);
- }
- /* pretend there's a final LF to close any blocks */
- handle_lf();
- return 0;
+ switch (c) {
+ case '\\':
+ escape = true;
+ break;
+ case '#':
+ if (in == NONE) {
+ in = HEADER_PARSE;
+ hlvl = 1;
+ } else if (in == HEADER_PARSE)
+ hlvl++;
+ else
+ putesc(c);
+ break;
+ case ' ':
+ if (in == HEADER_PARSE) {
+ printf("<h%d>", hlvl);
+ in = HEADER;
+ } else if (in == LIST_PARSE) {
+ if (!listdepth) {
+ newlist();
+ } else {
+ if (previndent < indent) {
+ newlist();
+ } else if (previndent > indent) {
+ endlist();
+ printf("</li>\n<li>\n");
+ } else {
+ in = LIST;
+ printf("</li>\n<li>\n");
+ }
+ }
+ } else
+ putesc(c);
+ break;
+ case '*':
+ case '`':
+ case '_':
+ toggle_format(c);
+ break;
+ case '\t':
+ if (in == NONE) {
+ in = CODE;
+ printf("<pre><code>");
+ } else if (lastc != '\n')
+ putesc(c);
+ break;
+ case '>':
+ if (in == NONE) {
+ in = PARAGRAPH;
+ printf("%s", blockquote ? "<p>\n" : "<blockquote>\n<p>\n");
+ blockquote = true;
+ } else if (lastc != '\n')
+ putesc(c);
+ break;
+ case '[':
+ if (in_link == NOL && !fmt_disabled(c)) {
+ maybe_startp();
+ in_link = DESC_PARSE;
+ lnkidx = 0;
+ } else
+ putesc(c);
+ break;
+ case ']':
+ if (in_link == DESC_PARSE)
+ in_link = OPT_URL;
+ else
+ putesc(c);
+ break;
+ case '(':
+ if (in_link == OPT_URL) {
+ in_link = URL_PARSE;
+ printf("<a href=\"");
+ } else
+ putesc(c);
+ break;
+ case ')':
+ if (in_link == URL_PARSE) {
+ in_link = NOL;
+ printf("\">");
+ for (int i = 0; i < lnkidx; i++)
+ handlec(lnkdes[i]);
+ printf("</a>");
+ } else
+ putesc(c);
+ break;
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ case '8':
+ case '9':
+ case '.':
+ case '-':
+ if (in == NONE || !linestarted) {
+ ol = c != '-';
+ in = LIST_PARSE;
+ } else if (in != LIST_PARSE)
+ putesc(c);
+ break;
+ case '\n':
+ handle_lf();
+ if (in != NONE)
+ putesc(c);
+ break;
+ default:
+ maybe_startp();
+ putesc(c);
+ break;
+ }
+ lastc = c;
+ linestarted = c != '\n';
}
-int main(int argc, char *argv[]) {
- if (argc > 1) {
- fprintf(stderr, "error: %s takes no arguments\n", argv[0]);
- return 1;
- }
+
+/* nihdoc: a text -> HTML parser */
+int
+main(int argc, char *argv[])
+{
+ int c;
+
#ifdef __OpenBSD__
- pledge("stdio", "stdio");
+ pledge("stdio", "stdio");
#endif
- return parse();
+
+ if (argc > 1) {
+ fprintf(stderr, "error: %s takes no arguments\n", argv[0]);
+ return 1;
+ }
+
+ while ((c = getchar()) != EOF)
+ handlec(c);
+
+ /* pretend there's a final LF to close any blocks */
+ handle_lf();
+
+ return 0;
}
