From 8386ee633abe6597903cc6439027ce1518847de2 Mon Sep 17 00:00:00 2001
From: Nicolas 'Pixel' Noble <nicolas@nobis-crew.org>
Date: Thu, 23 Jul 2020 09:51:52 -0700
Subject: [PATCH] Conversion script.

---
 conversion/psx-spx-pass1.awk | 137 +++++++++++++++++++++++++++++++++++
 conversion/run.sh            |  18 +++++
 2 files changed, 155 insertions(+)
 create mode 100644 conversion/psx-spx-pass1.awk
 create mode 100755 conversion/run.sh

diff --git a/conversion/psx-spx-pass1.awk b/conversion/psx-spx-pass1.awk
new file mode 100644
index 0000000..19529e7
--- /dev/null
+++ b/conversion/psx-spx-pass1.awk
@@ -0,0 +1,137 @@
+# Removes extra spaces off a string
+function ltrim(s) { sub(/^[ \t\r\n]+/, "", s); return s; }
+function rtrim(s) { sub(/[ \t\r\n]+$/, "", s); return s; }
+function trim(s) { return rtrim(ltrim(s)); }
+
+# Escape slashes and amps in a string for building links
+function linkescape(s) {
+    gsub(/\//, "\\/", s);
+    gsub(/\&/, "\\\\&", s);
+    return s;
+}
+
+# Creates a section link using normal replacement rules
+function nametolink(s) {
+    s = tolower(s);
+    gsub(/[^a-z0-9 -]/, "", s);
+    gsub(/ /, "-", s);
+    return s;
+}
+
+BEGIN {
+    sectionCount = 0;
+    WAIT_FOR_START = 1;
+    IN_PRE = 0;
+}
+
+# PARSE_HEADERS will be true only for the TOC list of HREFs
+# so we can build the list of "chapters".
+PARSE_HEADERS && /<A HREF=/ {
+    match($0, /"#(.*)">(.*)<.A/, m);
+    sections[sectionCount] = m[2];
+    sectionsLink[sectionCount] = m[1];
+    sectionCount++;
+    next;
+}
+
+# This marks the beginning of the TOC really.
+WAIT_FOR_START && /<B>Nocash PSXSPX Playstation Specifications<.B><BR>/ {
+    PARSE_HEADERS = 1;
+    next;
+}
+
+# These markers are to split off sections and subsections.
+/^<!--------------------------------------->/ {
+    if (PARSE_HEADERS) WAIT_FOR_START = 0;
+    PARSE_HEADERS = 0;
+    WAIT_SECTION = 1;
+    next;
+}
+
+# Trash anything that's in the TOC. We'll build a new one later.
+WAIT_FOR_START { next; }
+
+# We need to do slighly different parsing if we're in or out a PRE section.
+/<TABLE.*PRE>/ { IN_PRE = 1; }
+/<.TD><.TR><.TABLE>/ { IN_PRE = 0; }
+
+# Generic sed-like line replacements.
+{
+    # The input file isn't valid html to begin with,
+    # so we're not going to do a true html parser.
+    # Replace the typical html escapes right off the bat.
+    gsub(/&nbsp;/, " ");
+    if (IN_PRE) {
+        gsub(/&lt;/, "<");
+        gsub(/&gt;/, ">");
+    } else {
+        # outside of <PRE> blocks, we want to escape these for the md format.
+        gsub(/&lt;/, "\\<");
+        gsub(/&gt;/, "\\>");
+    }
+    gsub(/&amp;/, "\\&");
+
+    # These are fairly straightforward replacements
+    sub(/<TABLE.*PRE>/, "```");
+    sub(/<.TD><.TR><.TABLE>/, "```");
+    $0 = gensub(/<B>(.*)<.B><BR>/, "#### \\1", "g");
+    sub(/<BR>/, "<br/>");
+}
+
+# Skip the remaining cruft.
+/^<TABLE/ { next; }
+/^<.FONT/ { next; }
+
+# We need to parse the header of a major or minor section.
+WAIT_SECTION {
+    WAIT_SECTION = 0;
+    WAIT_HEADER = 1;
+    match($0, /<A NAME="(.*)">/, m);
+    link = m[1];
+    found = 0;
+    # Checking if it was in our TOC to distingish between section or subsection.
+    for (i = 0; i < sectionCount; i++) {
+        if (sectionsLink[i] == link) {
+            GOT_MAJOR_SECTION = 1;
+            sectionFile = link;
+            found = 1;
+            subsectionsCount = 0;
+            printf("SWITCHED TO MAJOR SECTION: %s\n", link);
+        }
+    }
+    if (!found) {
+        subsection = link;
+        printf("SWITCHED TO MINOR SECTION: %s\n", link);
+    }
+    next;
+}
+
+# Creating the proper section header, and the second pass sed script.
+WAIT_HEADER {
+    WAIT_HEADER = 0;
+    if (GOT_MAJOR_SECTION) {
+        print("s/<A HREF=\"#" link "\">[^<]*<.A>/[" linkescape(trim($0)) "](" link ".md)/g") > "psx-spx-pass2.sed";
+        $0 = "# " $0;
+    } else {
+        print("s/<A HREF=\"#" link "\">[^<]*<.A>/[" linkescape(trim($0)) "](" sectionFile ".md#" nametolink(trim($0)) ")/g") > "psx-spx-pass2.sed";
+        $0 = "## " $0;
+    }
+
+    GOT_MAJOR_SECTION = 0;
+}
+
+# If we get here, output our current line to the current section md file.
+{ print > sectionFile ".md" }
+
+END {
+    # Emit the TOC and finish the second pass sed script, as well as
+    # the second pass runner script.
+    for (i = 0; i < sectionCount; i++) {
+        link = sectionsLink[i];
+        name = sections[i];
+        print("sed -f psx-spx-pass2.sed -i " link ".md") > "run-pass2.sh";
+        print("[" name "](" link ".md)<br/>") > "index.md";
+    }
+
+    print("s/^<br\\/>$//") > "psx-spx-pass2.sed";
+}
diff --git a/conversion/run.sh b/conversion/run.sh
new file mode 100755
index 0000000..25c2b07
--- /dev/null
+++ b/conversion/run.sh
@@ -0,0 +1,18 @@
+#!/bin/sh
+
+# Download and slightly adjust the input file.
+# There's one occurence where there's a newline missing, which throws our awk script off.
+# Might as well fix it here rather than making the awk script more complex.
+curl https://problemkaputt.de/psx-spx.htm |
+  dos2unix |
+  sed 's|</FONT></TD></TR></TABLE><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR><TD><PRE>|</FONT></TD></TR></TABLE>\n<TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR><TD><PRE>|' > psx-spx.html
+
+# Run the awk script to generate all of the .md files, as well as the second pass scripts.
+gawk -f psx-spx-pass1.awk psx-spx.html
+
+# Invoke the second pass generated by the first pass script, to adjust all the cross references.
+. ./run-pass2.sh
+
+# Cleanup
+rm psx-spx.html run-pass2.sh psx-spx-pass2.sed
+mv *.md ..