switch yxml_ret_t to bitfield

HTML parsing may return multiple state change for 1 token
For example: ">" in "<input hidden/>" will return ATTRSTART | ELEMEND

.github/workflows/ci.yml

@@ -0,0 +1,32 @@
+name: CI
+
+on:
+  push:
+    branches: [ "master" ]
+  pull_request:
+    branches: [ "master" ]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+    - run: make html2json
+    - name: publish
+      env:
+        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        RELEASE_FILES: html2json
+      run: |
+        RELEASE_TAG="$(date +%y%m%d)"
+        curl -s -XPOST -d '{"tag_name": "'$RELEASE_TAG'"}' \
+          -H "Authorization: Bearer $GITHUB_TOKEN" \
+          -H 'Content-Type: application/json' \
+          "https://api.github.com/repos/${GITHUB_REPOSITORY}/releases" || :
+        RELEASE_ID=$(curl -s https://api.github.com/repos/${GITHUB_REPOSITORY}/releases/tags/$RELEASE_TAG | jq .id)
+        echo tag=$RELEASE_TAG has id=$RELEASE_ID
+        for RELEASE_FILE in $RELEASE_FILES; do
+          curl -s -XPOST -T $RELEASE_FILE \
+          -H "Authorization: token $GITHUB_TOKEN" \
+          -H "Content-Type:application/octet-stream" \
+          "https://uploads.github.com/repos/${GITHUB_REPOSITORY}/releases/$RELEASE_ID/assets?name=$RELEASE_FILE" || :
+        done

README.md

@@ -1,4 +1,5 @@
-Tiny HTML to JSON Converter
+Convert any XML/HTML to JsonML
+
 ## BUILD
 
 ```sh
@@ -37,17 +38,27 @@ cat test/basic.html | ./html2json | jq .[1].lang
 ```jsonc
 // doctype is ommited
 ["html",{"lang":"en"},[
-    ["head",{},[
+    ["head", {}, [
         ["meta", {"charset": "utf-8"} ],
-        ["title", {}, "Basic Example" ],
+        ["title", {}, ["Basic Example"] ],
         ["link", {"rel": "stylesheet"} ]
     ]],
     ["body", {"id": "home"}, [
-        ["input", {"type": "text"}]
-        ["p", {} ,"content"],
+        ["input", {"type": "text"}],
+        ["p", {}, ["content"]]
     ]]
 ]]
 ```
 
 </td></tr>
-</table>
+</table>
+
+# LIMITATIONS
+
+parsing is done by [yxml](https://dev.yorhel.nl/yxml) with the following changes for HTML support:
+- migrate `yxml_ret_t` to bitfield enum so multiple state can be returned (example : parsing `>` in `<p hidden>` will return `ATTREND|ELEMSTART`)
+- accept lowercase `<!doctype `
+- accept unquoted attribute value `<form method=GET>`
+- accept value-less attribute `<p hidden>`
+- (HTML5 mode) threat encoutered [void elements](https://developer.mozilla.org/en-US/docs/Glossary/Void_element) as self-closed
+- (HTML5 mode) ignore end-tag of void elements (ex: `</img>`)

html2json.c

@@ -20,68 +20,60 @@ int die(yxml_ret_t r, yxml_t* x)
     fprintf(stderr, "error %i at line %u col %u %s", r, x->line, x->byte, reason);
     return -r;
 }
-void putc_escape(int c)
+void putc_escape(char* c)
 {
-    if (c == '\\')
+    if (*c == '\\')
         printf("\\\\");
-    else if (c == '\r')
+    else if (*c == '\r')
         printf("\\r");
-    else if (c == '\n')
+    else if (*c == '\n')
         printf("\\n");
-    else if (c == '\t')
+    else if (*c == '\t')
         printf("\\t");
     else
-        printf("%c", c);
+        printf("%s", c);
 }
 int main(int argc, char** argv)
 {
     yxml_t x;
-    yxml_ret_t r;
+    yxml_ret_t old_r, r = YXML_OK;
     char buf[BUFSIZE];
+    setbuf(stdout, NULL);
     yxml_init(&x, buf, BUFSIZE);
-    size_t need_close_attr = 0, need_close_tag = 0, depth = 0, last_elem_depth = -1, need_sep_attr = 0, element_has_text = 0;
+    size_t depth = 0, last_elem_depth = -1, element_has_text = 0;
     for (int c; (c = getchar()) != EOF;) {
-        r = yxml_parse(&x, c);
-        if (0 && r)
-            printf("(%i)", r);
-        if (r < 0) {
-            return die(r, &x);
-        }
-        // pre-processing (autoclose,...)
-        if (need_close_attr && (r == YXML_ELEMSTART || r == YXML_CONTENT || r == YXML_ELEMEND)) {
-            printf("},[");
-            need_close_attr = 0;
+        yxml_ret_t tmp = yxml_parse(&x, c);
+        if (tmp < 0)
+            return die(tmp, &x);
+        if (tmp == YXML_OK)
+            continue;
+        element_has_text = tmp == YXML_CONTENT && ((c != ' ' && c != '\n' && c != '\r' && c != '\t') || element_has_text);
+        if (tmp == YXML_CONTENT && !element_has_text){
+            continue;
         }
+        old_r = r;
+        r = tmp;
 
-        if (r == YXML_ELEMSTART) {
-            if (last_elem_depth == depth)
-                printf(",");
-            printf("[\"%s\",{", x.elem);
-            depth++;
-            need_close_attr = 1;
-            need_sep_attr = 0;
-			element_has_text = 0;
-        } else if (r == YXML_ATTRSTART) {
-            if (need_sep_attr)
-                printf(",");
-            printf("\"%s\":\"", x.attr);
-            need_sep_attr = 1;
-        } else if (r == YXML_ATTRVAL) {
-            putc_escape(c);
-        } else if (r == YXML_ATTREND) {
-            printf("\"");
-        } else if (r == YXML_CONTENT) {
-            int is_space = (c == ' ' || c == '\n' || c == '\r' || c == '\t');
-			if(!element_has_text && !is_space){printf("\"");}
-            element_has_text |= !is_space;
-			if(element_has_text)putc_escape(c);
-        } else if (r == YXML_ELEMEND) {
-			if(element_has_text){printf("\"");}
-			element_has_text=0;
-            printf("]]");
-            depth--;
-            last_elem_depth = depth;
-        }
+        // global test, put it first
+        if (old_r & YXML_CONTENT && !(r & YXML_CONTENT)) printf("\"");
+
+        if (r & YXML_ELEMSTART && old_r & YXML_ELEMSTART)printf(", {}, [");
+        if (r & YXML_ELEMSTART) printf("%s", (last_elem_depth == depth || old_r & YXML_CONTENT)?",":"");//0,1,2,4
+        if (r & YXML_ELEMSTART) printf("\n%*.s[\"%s\"", 2*(depth++),"",x.elem);
+        if (r & YXML_ATTRSTART && old_r & YXML_ELEMSTART)printf(", {");
+        if (r & YXML_ATTRSTART) printf("%s\"%s\":",(old_r & YXML_ATTREND?",":""),x.attr);
+        if (r & YXML_ATTRVAL && old_r & YXML_ATTRSTART)printf("\"");
+        if (r & YXML_ATTRVAL) putc_escape(x.data);
+        if (r & YXML_ATTREND) printf("\"");
+        if (r & YXML_CONTENT && old_r & YXML_ATTREND) printf("}, [");
+        if (r & YXML_CONTENT && old_r & YXML_ELEMSTART) printf(", {}, [");// insert empty object for easier JQ
+        if (r & YXML_CONTENT && !(old_r & YXML_CONTENT)) printf("%s\"",last_elem_depth == depth ? ",":"");//1,2,4,32
+        if (r & YXML_CONTENT) putc_escape(x.data);
+        if (r & YXML_ELEMEND && (old_r & YXML_ELEMSTART)) printf(", {}, []");// insert empty child for easeir jq
+        if (r & YXML_ELEMEND && (old_r & YXML_ATTREND)) printf("}, []");// insert empty child for easeir jq
+        if (r & YXML_ELEMEND && (old_r & YXML_CONTENT)) printf("]");
+        if (r & YXML_ELEMEND && (old_r & YXML_ELEMEND)) printf("]");
+        if (r & YXML_ELEMEND) printf("]",2*(last_elem_depth = --depth),"");
     }
     r = yxml_eof(&x);
     if (r < 0)