Better article parsing, support for article code parsing (add line numbers to code blocks)

2 files changed, 64 insertions, 2 deletions

diff --git a/utils/parse-article-content.sh b/utils/parse-article-content.sh
new file mode 100755
index 0000000..40593c5
--- /dev/null
+++ b/utils/parse-article-content.sh
@@ -0,0 +1,62 @@
+#!/bin/bash
+
+# Skip file metadata, first 3 lines
+
+sanitize-html-entities () {
+  echo "$1" |
+  sed 's/</\&lt;/g' |
+  sed 's/>/\&gt;/g'
+}
+
+parse-article-content-file () {
+  local PRE_TAG=
+  local IGNORE_NEXT_PRE_TAG=
+  local INSIDE_PRE_TAG=
+  local CONTENT
+  local TLINE=
+
+  while IFS= read -r line; do
+    TLINE="$line"
+
+    if [ -z "$INSIDE_PRE_TAG" ]; then
+      echo "$line" | grep '<pre role="code"' > /dev/null
+      if [ $? -eq 0 ]; then
+        PRE_TAG=$line
+        INSIDE_PRE_TAG=1
+        TLINE="$line <!-- code-start -->"
+      fi
+    else
+      echo "$line" | grep '<pre' > /dev/null
+      if [ "$?" -eq 0 ]; then
+        IGNORE_NEXT_PRE_TAG=1
+        TLINE="<code>$(sanitize-html-entities "$line")</code>"
+      fi
+
+      echo "$line" | grep '</pre>' > /dev/null
+      if [ $? -eq 0 ]; then
+        if [ -n "$IGNORE_NEXT_PRE_TAG" ]; then
+          IGNORE_NEXT_PRE_TAG=""
+          TLINE="<code>$(sanitize-html-entities "$line")</code>"
+        else
+          INSIDE_PRE_TAG=""
+          PRE_TAG=""
+          TLINE="$line<!-- code-end -->"
+        fi
+      else
+        TLINE="<code>$(sanitize-html-entities "$line")</code>"
+      fi
+
+    fi
+
+    CONTENT="${CONTENT}
+$TLINE"
+  done <<< $(echo "$ARTICLE_FILE_CONTENT" | tail -n +3)
+
+  echo "$CONTENT"
+}
+
+if [ "$DO_NOT_PROCESS_HTML" = "true" ]; then
+  ARTICLE_CONTENT=$(echo "$ARTICLE_FILE_CONTENT" | tail -n +3)
+else
+  ARTICLE_CONTENT=$(parse-article-content-file)
+fi
diff --git a/utils/parse-article.sh b/utils/parse-article.sh
index c3e540a..b64c0ff 100755
--- a/utils/parse-article.sh
+++ b/utils/parse-article.sh
@@ -1,6 +1,7 @@
 #!/bin/bash
 
 ARTICLE_TITLE=$(basename "$ARTICLE_FILE")
+ARTICLE_FILE_SANITIZED="$(sanitize-filename "$ARTICLE_FILE")"
 
 # Pick all file content
 ARTICLE_FILE_CONTENT=$(cat "$ARTICLE_FILE")
@@ -8,8 +9,7 @@ ARTICLE_FILE_CONTENT=$(cat "$ARTICLE_FILE")
 # The first 3 lines of the file are metadata information
 ARTICLE_METADATA=$(echo "$ARTICLE_FILE_CONTENT" | head -n 3)
 
-# Skip file metadata, first 3 lines
-ARTICLE_CONTENT=$(echo "$ARTICLE_FILE_CONTENT" | tail -n +3)
+source $SCRIPT_DIR/utils/parse-article-content.sh
 
 # Tags are placed on the first line of the file, delimited by commas
 IFS=',' read -ra ARTICLE_TAGS <<< "$(echo "$ARTICLE_METADATA" | head -n 1)"