toml: fix parsing comments (denoland/deno_std#311)

Original: https://github.com/denoland/deno_std/commit/a48fdd6a298e48a60dd8d56f2b459294892abfc3

7 files changed, 64 insertions, 49 deletions

diff --git a/toml/parser.ts b/toml/parser.ts
index 9263f0111..d8acc608b 100644
--- a/toml/parser.ts
+++ b/toml/parser.ts
@@ -29,8 +29,9 @@ class Parser {
   _sanitize(): void {
     const out = [];
     for (let i = 0; i < this.tomlLines.length; i++) {
-      const s = this.tomlLines[i].split("#")[0];
-      if (s !== "") {
+      const s = this.tomlLines[i];
+      const trimmed = s.trim();
+      if (trimmed !== "" && trimmed[0] !== "#") {
         out.push(s);
       }
     }
@@ -156,7 +157,7 @@ class Parser {
   }
   _isGroup(line: string): boolean {
     const t = line.trim();
-    return t[0] === "[" && t[t.length - 1] === "]";
+    return t[0] === "[" && /\[(.*)\]/.exec(t) ? true : false;
   }
   _isDeclaration(line: string): boolean {
     return line.split("=").length > 1;
@@ -183,28 +184,50 @@ class Parser {
     kv.value = this._parseData(line.slice(idx + 1));
     return kv;
   }
+  // TODO (zekth) Need refactor using ACC
   _parseData(dataString: string): unknown {
     dataString = dataString.trim();
+
     if (this._isDate(dataString)) {
-      return new Date(dataString);
+      return new Date(dataString.split("#")[0].trim());
     }
+
     if (this._isLocalTime(dataString)) {
-      return eval(`"${dataString}"`);
+      return eval(`"${dataString.split("#")[0].trim()}"`);
     }
-    if (dataString === "inf" || dataString === "+inf") {
+
+    const cut3 = dataString.substring(0, 3).toLowerCase();
+    const cut4 = dataString.substring(0, 4).toLowerCase();
+    if (cut3 === "inf" || cut4 === "+inf") {
       return Infinity;
     }
-    if (dataString === "-inf") {
+    if (cut4 === "-inf") {
       return -Infinity;
     }
-    if (
-      dataString === "nan" ||
-      dataString === "+nan" ||
-      dataString === "-nan"
-    ) {
+
+    if (cut3 === "nan" || cut4 === "+nan" || cut4 === "-nan") {
       return NaN;
     }
-    // inline table
+
+    // If binary / octal / hex
+    const hex = /(0(?:x|o|b)[0-9a-f_]*)[^#]/gi.exec(dataString);
+    if (hex && hex[0]) {
+      return hex[0].trim();
+    }
+
+    const testNumber = this._isParsableNumber(dataString);
+    if (testNumber && !isNaN(testNumber as number)) {
+      return testNumber;
+    }
+
+    const invalidArr = /,\]/g.exec(dataString);
+    if (invalidArr) {
+      dataString = dataString.replace(/,]/g, "]");
+    }
+    const m = /(?:\'|\[|{|\").*(?:\'|\]|\"|})\s*[^#]/g.exec(dataString);
+    if (m) {
+      dataString = m[0].trim();
+    }
     if (dataString[0] === "{" && dataString[dataString.length - 1] === "}") {
       const reg = /([a-zA-Z0-9-_\.]*) (=)/gi;
       let result;
@@ -218,17 +241,6 @@ class Parser {
       // TODO : unflat if necessary
       return JSON.parse(dataString);
     }
-    // If binary / octal / hex
-    if (
-      dataString[0] === "0" &&
-      (dataString[1] === "b" || dataString[1] === "o" || dataString[1] === "x")
-    ) {
-      return dataString;
-    }
-
-    if (this._isParsableNumber(dataString)) {
-      return eval(dataString.replace(/_/g, ""));
-    }
 
     // Handle First and last EOL for multiline strings
     if (dataString.startsWith(`"\\n`)) {
@@ -241,18 +253,19 @@ class Parser {
     } else if (dataString.endsWith(`\\n'`)) {
       dataString = dataString.replace(`\\n'`, `'`);
     }
-
-    // dataString = dataString.replace(/\\/, "\\\\");
-
     return eval(dataString);
   }
   _isLocalTime(str: string): boolean {
     const reg = /(\d{2}):(\d{2}):(\d{2})/;
     return reg.test(str);
   }
-  _isParsableNumber(dataString: string): boolean {
-    let d = dataString.replace(/_/g, "");
-    return !isNaN(parseFloat(d));
+  _isParsableNumber(dataString: string): number | boolean {
+    const m = /((?:\+|-|)[0-9_\.e+\-]*)[^#]/i.exec(dataString.trim());
+    if (!m) {
+      return false;
+    } else {
+      return parseFloat(m[0].replace(/_/g, ""));
+    }
   }
   _isDate(dateStr: string): boolean {
     const reg = /\d{4}-\d{2}-\d{2}/;
diff --git a/toml/parser_test.ts b/toml/parser_test.ts
index cc6675296..104428b25 100644
--- a/toml/parser_test.ts
+++ b/toml/parser_test.ts
@@ -149,7 +149,8 @@ test({
       deno: "is",
       not: "[node]",
       regex: "<ic*s*>",
-      NANI: "何?!"
+      NANI: "何?!",
+      comment: "Comment inside # the comment"
     };
     const actual = parseFile(path.join(testFilesDir, "simple.toml"));
     assertEquals(actual, expected);
diff --git a/toml/testdata/boolean.toml b/toml/testdata/boolean.toml
index 92264888a..242d29c96 100644
--- a/toml/testdata/boolean.toml
+++ b/toml/testdata/boolean.toml
@@ -1,3 +1,3 @@
-[boolean]
+[boolean] # i hate comments
 bool1 = true
 bool2 = false
\ No newline at end of file
diff --git a/toml/testdata/cargo.toml b/toml/testdata/cargo.toml
index 5402d10a4..291aa7db6 100644
--- a/toml/testdata/cargo.toml
+++ b/toml/testdata/cargo.toml
@@ -53,4 +53,4 @@ tokio-threadpool = "0.1.11"
 url = "1.7.2"
 
 [target.'cfg(windows)'.dependencies]
-winapi = "0.3.6"
\ No newline at end of file
+winapi = "0.3.6"
diff --git a/toml/testdata/datetime.toml b/toml/testdata/datetime.toml
index b21924793..62377a4ba 100644
--- a/toml/testdata/datetime.toml
+++ b/toml/testdata/datetime.toml
@@ -1,8 +1,8 @@
 [datetime]
-odt1 = 1979-05-27T07:32:00Z
-odt2 = 1979-05-27T00:32:00-07:00
-odt3 = 1979-05-27T00:32:00.999999-07:00
-odt4 = 1979-05-27 07:32:00Z
-ld1 = 1979-05-27
-lt1 = 07:32:00
-lt2 = 00:32:00.999999
+odt1 = 1979-05-27T07:32:00Z # Comment
+odt2 = 1979-05-27T00:32:00-07:00 # Comment
+odt3 = 1979-05-27T00:32:00.999999-07:00 # Comment
+odt4 = 1979-05-27 07:32:00Z # Comment
+ld1 = 1979-05-27 # Comment
+lt1 = 07:32:00 # Comment
+lt2 = 00:32:00.999999 # Comment
diff --git a/toml/testdata/float.toml b/toml/testdata/float.toml
index 92f0b6d17..6a384179c 100644
--- a/toml/testdata/float.toml
+++ b/toml/testdata/float.toml
@@ -1,17 +1,17 @@
 [float]
 # fractional
-flt1 = +1.0
-flt2 = 3.1415
-flt3 = -0.01
+flt1 = +1.0 # Comment
+flt2 = 3.1415 # Comment
+flt3 = -0.01 # Comment
 
 # exponent
-flt4 = 5e+22
-flt5 = 1e6
-flt6 = -2E-2
+flt4 = 5e+22 # Comment
+flt5 = 1e6 # Comment
+flt6 = -2E-2 # Comment
 
 # both
-flt7 = 6.626e-34
-flt8 = 224_617.445_991_228
+flt7 = 6.626e-34 # Comment
+flt8 = 224_617.445_991_228 # Comment
 # infinity
 sf1 = inf  # positive infinity
 sf2 = +inf # positive infinity
diff --git a/toml/testdata/simple.toml b/toml/testdata/simple.toml
index aac44714e..f3f6c1036 100644
--- a/toml/testdata/simple.toml
+++ b/toml/testdata/simple.toml
@@ -1,4 +1,5 @@
 deno = "is"
 not = "[node]"
 regex = '<\i\c*\s*>'
-NANI = '何?!'
\ No newline at end of file
+NANI = '何?!'
+comment = "Comment inside # the comment" # Comment