@shapeshift-labs/frontier-lang-parser 0.3.34 → 0.3.36

package/README.md

@@ -230,11 +230,11 @@ import { inspectFrontierSourceSyntax } from '@shapeshift-labs/frontier-lang-pars
 const report = inspectFrontierSourceSyntax(source);
 
 if (report.summary.failClosed) {
-  console.error(report.summary.unknownKinds);
+  console.error(report.summary.unknownKinds, report.diagnostics);
 }
 ```
 
-Nested child syntax such as `render` blocks inside a `view` is not reported as an unknown top-level block; the parent parser owns those child rows. The report is evidence about parser coverage only. It keeps `autoMergeClaim` and `semanticEquivalenceClaim` false.
+Nested child syntax such as `render` blocks inside a `view` is not reported as an unknown top-level block; the parent parser owns those child rows. The report ignores braces inside quoted strings and comments when finding authored declaration boundaries, reports UTF-8 `sourceBytes`, and fails closed for malformed or unterminated blocks through `diagnostics`. The report is evidence about parser coverage only. It keeps `autoMergeClaim` and `semanticEquivalenceClaim` false.
 
 ## Authored view render graph syntax
 

package/dist/index.d.ts

@@ -1,6 +1,11 @@
 import type { FrontierLangDocument } from '@shapeshift-labs/frontier-lang-kernel';
 export interface ParseFrontierOptions { readonly id?: string; readonly name?: string; }
 export declare const FrontierSourceBlockKinds: readonly string[];
+export interface FrontierSourceSyntaxDiagnostic {
+  readonly reason: 'unterminated-block' | 'unmatched-close-brace';
+  readonly message: string;
+  readonly location: { readonly line: number; readonly column: number; readonly offset: number };
+}
 export interface FrontierSourceBlockSyntaxRecord {
   readonly kind: string;
   readonly name: string;
@@ -14,6 +19,8 @@ export interface FrontierSourceBlockSyntaxRecord {
   readonly moduleId?: string;
   readonly moduleName?: string;
   readonly recognized: boolean;
+  readonly malformed?: boolean;
+  readonly diagnostics?: readonly FrontierSourceSyntaxDiagnostic[];
 }
 export interface FrontierUnknownSourceBlockSyntaxRecord extends FrontierSourceBlockSyntaxRecord {
   readonly recognized: false;
@@ -31,11 +38,14 @@ export interface FrontierSourceSyntaxReport {
     readonly blockCount: number;
     readonly recognizedBlockCount: number;
     readonly unknownBlockCount: number;
+    readonly malformedBlockCount: number;
+    readonly diagnosticCount: number;
     readonly recognizedKinds: readonly string[];
     readonly unknownKinds: readonly string[];
     readonly failClosed: boolean;
     readonly unsupportedSyntax: boolean;
   };
+  readonly diagnostics: readonly FrontierSourceSyntaxDiagnostic[];
   readonly metadata: {
     readonly sourceBytes: number;
     readonly autoMergeClaim: false;

package/dist/index.js

@@ -16,7 +16,7 @@ import { parseRuntimeCapabilityBlock } from './runtime-capability.js';
 import { parseNativeSourceBlock } from './source-evidence.js';
 import { parseTargetProjectionMetadata } from './target-projection.js';
 import { parseViewBlock } from './view.js';
-import { FrontierSourceBlockKinds } from './source-syntax-report.js';
+import { FrontierSourceBlockKinds, readFrontierSourceBlocks } from './source-syntax-report.js';
 export { FrontierSourceBlockKinds, inspectFrontierSourceSyntax } from './source-syntax-report.js';
 
 export function parseFrontierSource(source, options = {}) {
@@ -77,16 +77,7 @@ export function parseFrontierFile(name, source) { return parseFrontierSource(sou
 function readName(source) { return /module\s+([A-Za-z_$][\w$]*)/.exec(source)?.[1]; }
 function readId(source) { return /module\s+[A-Za-z_$][\w$]*\s+@id\(\s*["']([^"']+)["']\s*\)/.exec(source)?.[1]; }
 function readBlocks(source) {
-  const blocks = [];
-  const header = new RegExp('\\b(' + FrontierSourceBlockKinds.join('|') + ')\\s+([^{}]+)\\{', 'g');
-  let match;
-  while ((match = header.exec(source))) {
-    let depth = 1; let index = header.lastIndex;
-    while (index < source.length && depth > 0) { const ch = source[index++]; if (ch === '{') depth++; if (ch === '}') depth--; }
-    blocks.push({ kind: match[1], header: match[2].trim(), body: source.slice(header.lastIndex, index - 1) });
-    header.lastIndex = index;
-  }
-  return blocks;
+  return readFrontierSourceBlocks(source);
 }
 function idFrom(header, fallback) { return /@id\(\s*["']([^"']+)["']\s*\)/.exec(header)?.[1] ?? fallback; }
 function nameFrom(header) { return /^([A-Za-z_$][\w$]*)/.exec(header)?.[1] ?? 'Unnamed'; }

package/dist/source-syntax-report.js

@@ -48,7 +48,8 @@ const FrontierSourceBlockKindSet = new Set(FrontierSourceBlockKinds);
 export function inspectFrontierSourceSyntax(source, options = {}) {
   const documentId = options.id ?? readId(source) ?? 'mod_frontier';
   const documentName = options.name ?? readName(source) ?? 'FrontierModule';
-  const blocks = readCandidateDeclarationBlocks(source).map((block) => ({
+  const structure = scanFrontierStructure(source);
+  const blocks = readCandidateDeclarationBlocks(source, structure).map((block) => ({
     ...block,
     recognized: FrontierSourceBlockKindSet.has(block.kind)
   }));
@@ -57,6 +58,22 @@ export function inspectFrontierSourceSyntax(source, options = {}) {
     ...block,
     reason: 'unsupported-top-level-block'
   }));
+  const malformedBlockOpenOffsets = new Set(blocks.filter((block) => block.malformed).map((block) => block.bodyStartOffset - 1));
+  const diagnostics = [
+    ...blocks.flatMap((block) => block.diagnostics ?? []),
+    ...structure.unmatchedOpenBraces.filter((offset) => !malformedBlockOpenOffsets.has(offset)).map((offset) => ({
+      reason: 'unterminated-block',
+      message: 'Found an opening brace without a matching closing brace.',
+      location: sourcePosition(source, offset)
+    })),
+    ...structure.unmatchedCloseBraces.map((offset) => ({
+      reason: 'unmatched-close-brace',
+      message: 'Found a closing brace without a matching opening brace.',
+      location: sourcePosition(source, offset)
+    }))
+  ];
+  const malformedBlocks = blocks.filter((block) => block.malformed);
+  const failClosed = unknownBlocks.length > 0 || diagnostics.length > 0;
   return {
     kind: 'frontier.lang.sourceSyntaxReport',
     version: 1,
@@ -69,21 +86,57 @@ export function inspectFrontierSourceSyntax(source, options = {}) {
       blockCount: blocks.length,
       recognizedBlockCount: recognizedBlocks.length,
       unknownBlockCount: unknownBlocks.length,
+      malformedBlockCount: malformedBlocks.length,
+      diagnosticCount: diagnostics.length,
       recognizedKinds: unique(recognizedBlocks.map((block) => block.kind)),
       unknownKinds: unique(unknownBlocks.map((block) => block.kind)),
-      failClosed: unknownBlocks.length > 0,
+      failClosed,
       unsupportedSyntax: unknownBlocks.length > 0
     },
+    diagnostics,
     metadata: {
-      sourceBytes: source.length,
+      sourceBytes: utf8ByteLength(source),
       autoMergeClaim: false,
       semanticEquivalenceClaim: false
     }
   };
 }
 
-function readCandidateDeclarationBlocks(source) {
-  const moduleRanges = readModuleRanges(source);
+export function readFrontierSourceBlocks(source, options = {}) {
+  const report = inspectFrontierSourceSyntax(source, options);
+  return report.recognizedBlocks.filter((block) => !block.malformed).map((block) => ({
+    kind: block.kind,
+    header: block.header,
+    body: source.slice(block.bodyStartOffset, block.bodyEndOffset),
+    syntax: block
+  }));
+}
+
+export function readFrontierNestedBlocks(kind, source) {
+  const structure = scanFrontierStructure(source);
+  const blocks = [];
+  const header = new RegExp('\\b' + escapeRegExp(kind) + '\\s+([^{}]+?)\\{', 'g');
+  let match;
+  while ((match = header.exec(source))) {
+    const start = match.index;
+    const open = header.lastIndex - 1;
+    if (!structure.codeOffsets[start] || !structure.codeOffsets[open]) continue;
+    const close = findMatchingBrace(structure, open);
+    if (close < 0) continue;
+    blocks.push({
+      kind,
+      header: match[1].trim(),
+      body: source.slice(header.lastIndex, close),
+      start,
+      end: close + 1
+    });
+    header.lastIndex = close + 1;
+  }
+  return blocks;
+}
+
+function readCandidateDeclarationBlocks(source, structure) {
+  const moduleRanges = readModuleRanges(source, structure);
   const blocks = [];
   const header = /(^|\n)\s*([A-Za-z_$][\w$]*)\s+([^{}\n]+)\{/g;
   let match;
@@ -94,8 +147,12 @@ function readCandidateDeclarationBlocks(source) {
     const kind = match[2];
     if (kind === 'module') continue;
     const open = header.lastIndex - 1;
-    const close = findMatchingBrace(source, open);
-    const depth = braceDepthBefore(source, start);
+    if (!structure.codeOffsets[start] || !structure.codeOffsets[open]) continue;
+    const close = findMatchingBrace(structure, open);
+    const malformed = close < 0;
+    const end = malformed ? source.length : close + 1;
+    const bodyEnd = malformed ? source.length : close;
+    const depth = braceDepthBefore(structure, start);
     const moduleRange = moduleRanges.find((range) => start > range.open && start < range.close);
     const declarationDepth = moduleRange ? moduleRange.depth + 1 : 0;
     if (depth !== declarationDepth) continue;
@@ -106,18 +163,24 @@ function readCandidateDeclarationBlocks(source) {
       id: idFrom(headerText),
       header: headerText,
       startOffset: start,
-      endOffset: close + 1,
+      endOffset: end,
       bodyStartOffset: open + 1,
-      bodyEndOffset: close,
+      bodyEndOffset: bodyEnd,
       location: sourcePosition(source, start),
       moduleId: moduleRange?.id,
-      moduleName: moduleRange?.name
+      moduleName: moduleRange?.name,
+      malformed,
+      diagnostics: malformed ? [{
+        reason: 'unterminated-block',
+        message: `Block "${kind}" has no matching closing brace.`,
+        location: sourcePosition(source, open)
+      }] : []
     });
   }
   return blocks;
 }
 
-function readModuleRanges(source) {
+function readModuleRanges(source, structure) {
   const ranges = [];
   const header = /(^|\n)\s*module\s+([^{}\n]+)\{/g;
   let match;
@@ -126,36 +189,120 @@ function readModuleRanges(source) {
     const leading = /^\s*/.exec(source.slice(fullStart))?.[0].length ?? 0;
     const start = fullStart + leading;
     const open = header.lastIndex - 1;
-    const close = findMatchingBrace(source, open);
+    if (!structure.codeOffsets[start] || !structure.codeOffsets[open]) continue;
+    const close = findMatchingBrace(structure, open);
+    const malformed = close < 0;
     ranges.push({
       start,
       open,
-      close,
-      depth: braceDepthBefore(source, start),
+      close: malformed ? source.length : close,
+      depth: braceDepthBefore(structure, start),
       name: nameFrom(match[2].trim()),
-      id: idFrom(match[2].trim())
+      id: idFrom(match[2].trim()),
+      malformed
     });
   }
   return ranges;
 }
 
-function findMatchingBrace(source, open) {
-  let depth = 1;
-  for (let index = open + 1; index < source.length; index++) {
-    if (source[index] === '{') depth++;
-    if (source[index] === '}') depth--;
-    if (depth === 0) return index;
-  }
-  return source.length - 1;
+function findMatchingBrace(structure, open) {
+  return structure.bracePairs.get(open) ?? -1;
 }
 
-function braceDepthBefore(source, offset) {
+function braceDepthBefore(structure, offset) {
+  return structure.depthBefore[Math.min(Math.max(offset, 0), structure.depthBefore.length - 1)] ?? 0;
+}
+
+function scanFrontierStructure(source) {
+  const depthBefore = new Int32Array(source.length + 1);
+  const codeOffsets = new Uint8Array(source.length);
+  const stack = [];
+  const bracePairs = new Map();
+  const unmatchedCloseBraces = [];
   let depth = 0;
-  for (let index = 0; index < offset; index++) {
-    if (source[index] === '{') depth++;
-    if (source[index] === '}') depth = Math.max(0, depth - 1);
+  let state = 'code';
+  let quote = '';
+  for (let index = 0; index < source.length; index++) {
+    depthBefore[index] = depth;
+    const char = source[index];
+    const next = source[index + 1];
+    if (state === 'line-comment') {
+      if (char === '\n') state = 'code';
+      continue;
+    }
+    if (state === 'block-comment') {
+      if (char === '*' && next === '/') {
+        depthBefore[index + 1] = depth;
+        index++;
+        state = 'code';
+      }
+      continue;
+    }
+    if (state === 'string') {
+      if (char === '\\') {
+        depthBefore[index + 1] = depth;
+        index++;
+        continue;
+      }
+      if (char === quote) {
+        state = 'code';
+        quote = '';
+      }
+      continue;
+    }
+    codeOffsets[index] = 1;
+    if (char === '/' && next === '/') {
+      codeOffsets[index + 1] = 0;
+      depthBefore[index + 1] = depth;
+      index++;
+      state = 'line-comment';
+      continue;
+    }
+    if (char === '/' && next === '*') {
+      codeOffsets[index + 1] = 0;
+      depthBefore[index + 1] = depth;
+      index++;
+      state = 'block-comment';
+      continue;
+    }
+    if (char === '#' && isLineLeadingWhitespace(source, index)) {
+      state = 'line-comment';
+      continue;
+    }
+    if (char === '"' || char === "'" || char === '`') {
+      state = 'string';
+      quote = char;
+      continue;
+    }
+    if (char === '{') {
+      stack.push(index);
+      depth++;
+      continue;
+    }
+    if (char === '}') {
+      const open = stack.pop();
+      if (open === undefined) {
+        unmatchedCloseBraces.push(index);
+        continue;
+      }
+      depth = Math.max(0, depth - 1);
+      bracePairs.set(open, index);
+    }
+  }
+  depthBefore[source.length] = depth;
+  return { bracePairs, codeOffsets, depthBefore, unmatchedCloseBraces, unmatchedOpenBraces: stack };
+}
+
+function isLineLeadingWhitespace(source, offset) {
+  for (let index = offset - 1; index >= 0 && source[index] !== '\n'; index--) {
+    if (!/\s/.test(source[index])) return false;
   }
-  return depth;
+  return true;
+}
+
+function utf8ByteLength(source) {
+  if (typeof TextEncoder !== 'undefined') return new TextEncoder().encode(source).length;
+  return unescape(encodeURIComponent(source)).length;
 }
 
 function sourcePosition(source, offset) {
@@ -168,3 +315,4 @@ function readId(source) { return /module\s+[A-Za-z_$][\w$]*\s+@id\(\s*["']([^"']
 function idFrom(header) { return /@id\(\s*["']([^"']+)["']\s*\)/.exec(header)?.[1]; }
 function nameFrom(header) { return /^([A-Za-z_$][\w$]*)/.exec(header)?.[1] ?? 'Unnamed'; }
 function unique(values) { return [...new Set(values.filter(Boolean))]; }
+function escapeRegExp(value) { return String(value).replace(/[.*+?^${}()|[\]\\]/g, '\\$&'); }

package/dist/view.js

@@ -1,4 +1,5 @@
 import { viewNode } from '@shapeshift-labs/frontier-lang-kernel';
+import { readFrontierNestedBlocks } from './source-syntax-report.js';
 
 export function parseViewBlock(block) {
   const name = nameFrom(block.header);
@@ -79,16 +80,7 @@ function readRenderEvents(body) {
 }
 
 function readNestedBlocks(kind, source) {
-  const blocks = [];
-  const header = new RegExp('\\b' + kind + '\\s+([^{}]+)\\{', 'g');
-  let match;
-  while ((match = header.exec(source))) {
-    let depth = 1; let index = header.lastIndex;
-    while (index < source.length && depth > 0) { const ch = source[index++]; if (ch === '{') depth++; if (ch === '}') depth--; }
-    blocks.push({ kind, header: match[1].trim(), body: source.slice(header.lastIndex, index - 1), start: match.index, end: index });
-    header.lastIndex = index;
-  }
-  return blocks;
+  return readFrontierNestedBlocks(kind, source);
 }
 
 function stripNestedBlocks(kind, source) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@shapeshift-labs/frontier-lang-parser",
-  "version": "0.3.34",
+  "version": "0.3.36",
   "description": "Parser for the first Frontier Lang .frontier syntax slice.",
   "type": "module",
   "main": "./dist/index.js",