Use pup for HTML parsing; index the index pages

3 files changed, 82 insertions, 34 deletions

diff --git a/Makefile b/Makefile
index 344a120..397d539 100644
--- a/Makefile
+++ b/Makefile
@@ -50,7 +50,15 @@ $(INFO_PLIST_FILE): src/Info.plist $(CONTENTS_DIR)
 $(INDEX_FILE): src/index.sh $(DOCUMENTS_DIR)
 	rm -f $@
 	src/index.sh $@ $(DOCUMENTS_DIR)/*.html
-	#ruby src/index.rb $(DOCUMENTS_DIR)/*.html | sqlite3 $@
+	src/index.sh -i "Entry" $@ $(DOCUMENTS_DIR)/Concept-Index.html
+	src/index.sh -i "Macro" $@ $(DOCUMENTS_DIR)/M4-Macro-Index.html
+	src/index.sh -i "Macro" $@ $(DOCUMENTS_DIR)/Autoconf-Macro-Index.html
+	src/index.sh -i "Macro" $@ $(DOCUMENTS_DIR)/Autotest-Macro-Index.html
+	src/index.sh -i "Variable" $@ $(DOCUMENTS_DIR)/Cache-Variable-Index.html
+	src/index.sh -i "Variable" $@ $(DOCUMENTS_DIR)/Output-Variable-Index.html
+	src/index.sh -i "Function" $@ $(DOCUMENTS_DIR)/Program-_0026-Function-Index.html
+	src/index.sh -i "Entry" $@ $(DOCUMENTS_DIR)/Preprocessor-Symbol-Index.html
+	src/index.sh -i "Variable" $@ $(DOCUMENTS_DIR)/Environment-Variable-Index.html
 
 $(ICON_FILE): src/icon.png $(DOCSET_DIR)
 	cp src/icon.png $@
diff --git a/README b/README
index 1e64579..09134fb 100644
--- a/README
+++ b/README
@@ -12,4 +12,5 @@ Requirements:
 - any POSIX-compliant shell
 - curl
 - make
+- pup
 - sqlite3
diff --git a/src/index.sh b/src/index.sh
index e81a0f7..8132287 100755
--- a/src/index.sh
+++ b/src/index.sh
@@ -1,25 +1,16 @@
 #!/usr/bin/env sh
 
-DB_PATH="$1"
-shift
+create_table() {
+	sqlite3 "$DB_PATH" "CREATE TABLE IF NOT EXISTS searchIndex(id INTEGER PRIMARY KEY, name TEXT, type TEXT, path TEXT);"
+	sqlite3 "$DB_PATH" "CREATE UNIQUE INDEX IF NOT EXISTS anchor ON searchIndex (name, type, path);"
+}
 
 get_title() {
 	FILE="$1"
 
-	PATTERN="<title>.*\(Autoconf\).*</title>"
-
-	#Find pattern in file
-	grep -Eo "$PATTERN" "$FILE" | 
-		#Remove tag
-		sed 's/<[^>]*>//g' | \
-		#Remove '(automake)'
+	pup -p -f "$FILE" 'title text{}' | \
 		sed 's/(Autoconf)//g' | \
-		#Remove trailing space
-		sed 's/[ ]*$//g' | \
-		#Replace '&amp' with '&'
-		sed 's/&amp/&/g' | \
-		#Replace '&lt;' with '<'
-		sed 's/&lt;/</g'
+		sed 's/\"/\"\"/g'
 }
 
 get_type() {
@@ -44,23 +35,71 @@ insert() {
 	sqlite3 "$DB_PATH" "INSERT INTO searchIndex(name, type, path) VALUES (\"$NAME\",\"$TYPE\",\"$PAGE_PATH\");"
 }
 
-# Create table
-sqlite3 "$DB_PATH" "CREATE TABLE searchIndex(id INTEGER PRIMARY KEY, name TEXT, type TEXT, path TEXT);"
-sqlite3 "$DB_PATH" "CREATE UNIQUE INDEX anchor ON searchIndex (name, type, path);"
-
-# Get title and insert into table for each html file
-while [ -n "$1" ]; do
-	unset PAGE_NAME
-	unset PAGE_TYPE
-	PAGE_NAME="$(get_title "$1")"
-	if [ -n "$PAGE_NAME" ]; then
-		PAGE_TYPE="$(get_type "$1")"
-		#get_type "$1"
-		if [ -z "$PAGE_TYPE" ]; then
-			PAGE_TYPE="Guide"
+insert_index_terms() {
+	# Get each term from an index page and insert
+	while [ -n "$1" ]; do
+		grep -Eo "<a href.*></a>" "$1" | while read -r line; do
+			insert_term "$line"
+		done
+
+		shift
+	done
+}
+
+
+insert_pages() {
+	# Get title and insert into table for each html file
+	while [ -n "$1" ]; do
+		unset PAGE_NAME
+		unset PAGE_TYPE
+		PAGE_NAME="$(get_title "$1")"
+		if [ -n "$PAGE_NAME" ]; then
+			PAGE_TYPE="$(get_type "$1")"
+			#get_type "$1"
+			if [ -z "$PAGE_TYPE" ]; then
+				PAGE_TYPE="Guide"
+			fi
+			#echo "$PAGE_TYPE"
+			insert "$PAGE_NAME" "$PAGE_TYPE" "$(basename "$1")"
 		fi
-		#echo "$PAGE_TYPE"
-		insert "$PAGE_NAME" "$PAGE_TYPE" "$(basename "$1")"
-	fi
-	shift
+		shift
+	done
+}
+
+insert_term() {
+	LINK="$1"
+	NAME="$(echo "$LINK" | pup -p 'a text{}' | sed 's/"/\"\"/g')"
+	TYPE="$INDEX_TYPE"
+	PAGE_PATH="$(echo "$LINK" | pup -p 'a attr{href}')"
+
+	insert "$NAME" "$TYPE" "$PAGE_PATH"
+}
+
+TYPE="PAGES"
+
+# Check flags
+while true; do
+	case "$1" in
+		-i|--index)
+			TYPE="INDEX"
+			shift
+			INDEX_TYPE="$1"
+			shift
+			;;
+		*)
+			break
+	esac
 done
+
+DB_PATH="$1"
+shift
+
+create_table
+case "$TYPE" in
+	PAGES)
+		insert_pages "$@"
+		;;
+	INDEX)
+		insert_index_terms "$@"
+		;;
+esac