From df73a3baa016a55a46d248516281054cc45e03b3 Mon Sep 17 00:00:00 2001
From: David Chen <davidchen@gliacloud.com>
Date: Mon, 22 Jul 2024 20:21:25 -0700
Subject: [PATCH] fix #134 hotfix decode error (#135)

* fix #134 hotfix decode error

* fix pre-commit

* update

* loose version
---
 .pre-commit-config.yaml                       | 23 ++++++-------------
 poetry.lock                                   | 15 ++++++++++--
 pyproject.toml                                |  1 +
 src/fuzzy_json/decoder.py                     | 15 +++++++++---
 ...paired_json_invaild_case[case9.jsonx].json |  4 ++++
 .../tests/test_data/invalid/case9.jsonx       |  1 +
 6 files changed, 38 insertions(+), 21 deletions(-)
 create mode 100644 src/fuzzy_json/tests/__snapshots__/test_decoder/test_repaired_json_invaild_case[case9.jsonx].json
 create mode 100644 src/fuzzy_json/tests/test_data/invalid/case9.jsonx

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index bbdffd0..f560ae8 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -45,24 +45,15 @@ repos:
     -   id: codespell
         types_or: [python, markdown]
 
-  - repo: local
+  - repo: https://github.com/pre-commit/mirrors-mypy
+    rev: v1.8.0
     hooks:
       - id: mypy
         name: mypy
         entry: mypy
-        language: system
         types: [python]
-        exclude: migrations/|commands/|scripts/
-        args:
-          [
-            --pretty,
-            --show-error-codes,
-            --implicit-optional,
-            --follow-imports=silent,
-            --warn-redundant-casts,
-            --warn-unused-ignores,
-            --disallow-any-generics,
-            --check-untyped-defs,
-            --no-implicit-reexport,
-            --disallow-untyped-defs,
-          ]
+        exclude: migrations/|commands/|sandbox/|samples|sdk
+        additional_dependencies: [pytest, syrupy, json5]
+        args: [
+          "--config-file=pyproject.toml"
+        ]
diff --git a/poetry.lock b/poetry.lock
index 91a14e4..3e964e7 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -1,4 +1,4 @@
-# This file is automatically @generated by Poetry 1.8.3 and should not be changed by hand.
+# This file is automatically @generated by Poetry 1.8.2 and should not be changed by hand.
 
 [[package]]
 name = "certifi"
@@ -303,6 +303,17 @@ files = [
     {file = "iniconfig-2.0.0.tar.gz", hash = "sha256:2d91e135bf72d31a410b17c16da610a82cb55f6b0477d1a902134b24a455b8b3"},
 ]
 
+[[package]]
+name = "json5"
+version = "0.9.25"
+description = "A Python implementation of the JSON5 data format."
+optional = false
+python-versions = ">=3.8"
+files = [
+    {file = "json5-0.9.25-py3-none-any.whl", hash = "sha256:34ed7d834b1341a86987ed52f3f76cd8ee184394906b6e22a1e0deb9ab294e8f"},
+    {file = "json5-0.9.25.tar.gz", hash = "sha256:548e41b9be043f9426776f05df8635a00fe06104ea51ed24b67f908856e151ae"},
+]
+
 [[package]]
 name = "multidict"
 version = "6.0.5"
@@ -945,4 +956,4 @@ multidict = ">=4.0"
 [metadata]
 lock-version = "2.0"
 python-versions = "^3.10"
-content-hash = "bc2d9cdfa150d9633d0b33ebb02f08278bd84fbf195d4b7e1001c33ab3f6a519"
+content-hash = "a763de743853f83831433113d92fdc43d2f2cf6cbde655146928653670b3f7cf"
diff --git a/pyproject.toml b/pyproject.toml
index 47a42cb..3e80397 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -24,6 +24,7 @@ classifiers = [
 
 [tool.poetry.dependencies]
 python = "^3.10"
+json5 = "*"
 
 [tool.poetry.group.dev]
 optional = true
diff --git a/src/fuzzy_json/decoder.py b/src/fuzzy_json/decoder.py
index 868b686..0373442 100644
--- a/src/fuzzy_json/decoder.py
+++ b/src/fuzzy_json/decoder.py
@@ -2,6 +2,8 @@
 from functools import wraps
 from typing import Any, Callable
 
+import json5
+
 
 def state(fn: Callable[[str, list[str]], str | None]) -> Callable[[str, list[str]], str]:
     @wraps(fn)
@@ -288,10 +290,17 @@ def repair_json(json_str: str) -> str:
     return state_start(json_str)
 
 
-def loads(json_str: str, auto_repair: bool = True) -> dict[str, Any]:
+def base_loads(json_str: str) -> dict[str, Any]:
     try:
+        return json5.loads(json_str)
+    except Exception:
         return json.loads(json_str, strict=False)
-    except json.decoder.JSONDecodeError:
+
+
+def loads(json_str: str, auto_repair: bool = True) -> dict[str, Any]:
+    try:
+        return base_loads(json_str)
+    except Exception:
         if not auto_repair:
             raise
 
@@ -300,4 +309,4 @@ def loads(json_str: str, auto_repair: bool = True) -> dict[str, Any]:
     except Exception as e:
         raise json.decoder.JSONDecodeError(f"Failed to repair JSON: {e}", json_str, 0)
 
-    return json.loads(repaired_json, strict=False)
+    return base_loads(repaired_json)
diff --git a/src/fuzzy_json/tests/__snapshots__/test_decoder/test_repaired_json_invaild_case[case9.jsonx].json b/src/fuzzy_json/tests/__snapshots__/test_decoder/test_repaired_json_invaild_case[case9.jsonx].json
new file mode 100644
index 0000000..15def5b
--- /dev/null
+++ b/src/fuzzy_json/tests/__snapshots__/test_decoder/test_repaired_json_invaild_case[case9.jsonx].json
@@ -0,0 +1,4 @@
+{
+  "highlighted_sentence": "to choose \"a candidate who can <em>defeat Donald Trump</em> in November.\"",
+  "keyword": "defeat Donald Trump"
+}
diff --git a/src/fuzzy_json/tests/test_data/invalid/case9.jsonx b/src/fuzzy_json/tests/test_data/invalid/case9.jsonx
new file mode 100644
index 0000000..7a81131
--- /dev/null
+++ b/src/fuzzy_json/tests/test_data/invalid/case9.jsonx
@@ -0,0 +1 @@
+{"keyword": "defeat Donald Trump", "highlighted_sentence": 'to choose "a candidate who can <em>defeat Donald Trump</em> in November."'}