json-repair 0.28.4__py3-none-any.whl → 0.29.0__py3-none-any.whl

json_repair/__main__.py

@@ -0,0 +1,4 @@
+from .json_repair import cli
+
+if __name__ == "__main__":
+    cli()

json_repair/json_repair.py

@@ -22,7 +22,9 @@ If something is wrong (a missing parantheses or quotes for example) it will use
 All supported use cases are in the unit tests
 """
 
+import argparse
 import os
+import sys
 import json
 from typing import Any, Dict, List, Optional, Union, TextIO, Tuple, Literal
 
@@ -40,14 +42,16 @@ class StringFileWrapper:
             CHUNK_LENGTH = 1_000_000
         self.buffer_length = CHUNK_LENGTH
 
-    def fill_buffer(self, index: int) -> None:
+    def get_buffer(self, index: int) -> str:
         if self.buffers.get(index) is None:
             self.fd.seek(index * self.buffer_length)
             self.buffers[index] = self.fd.read(self.buffer_length)
             # Save memory by keeping max 2MB buffer chunks and min 2 chunks
             if len(self.buffers) > max(2, 2_000_000 / self.buffer_length):
                 oldest_key = next(iter(self.buffers))
-                self.buffers.pop(oldest_key)
+                if oldest_key != index:
+                    self.buffers.pop(oldest_key)
+        return self.buffers[index]
 
     def __getitem__(self, index: Union[int, slice]) -> str:
         # The buffer is an array that is seek like a RAM:
@@ -56,25 +60,24 @@ class StringFileWrapper:
         if isinstance(index, slice):
             buffer_index = index.start // self.buffer_length
             buffer_end = index.stop // self.buffer_length
-            for i in range(buffer_index, buffer_end + 1):
-                self.fill_buffer(i)
             if buffer_index == buffer_end:
-                return self.buffers[buffer_index][
+                return self.get_buffer(buffer_index)[
                     index.start % self.buffer_length : index.stop % self.buffer_length
                 ]
             else:
-                start_slice = self.buffers[buffer_index][
+                start_slice = self.get_buffer(buffer_index)[
                     index.start % self.buffer_length :
                 ]
-                end_slice = self.buffers[buffer_end][: index.stop % self.buffer_length]
+                end_slice = self.get_buffer(buffer_end)[
+                    : index.stop % self.buffer_length
+                ]
                 middle_slices = [
-                    self.buffers[i] for i in range(buffer_index + 1, buffer_end)
+                    self.get_buffer(i) for i in range(buffer_index + 1, buffer_end)
                 ]
                 return start_slice + "".join(middle_slices) + end_slice
         else:
             buffer_index = index // self.buffer_length
-            self.fill_buffer(buffer_index)
-            return self.buffers[buffer_index][index % self.buffer_length]
+            return self.get_buffer(buffer_index)[index % self.buffer_length]
 
     def __len__(self) -> int:
         if self.length < 1:
@@ -755,3 +758,47 @@ def from_file(
     fd.close()
 
     return jsonobj
+
+
+def cli():  # pragma: no cover
+    parser = argparse.ArgumentParser(description="Repair and parse JSON files.")
+    parser.add_argument("filename", help="The JSON file to repair")
+    parser.add_argument(
+        "-i",
+        "--inline",
+        action="store_true",
+        help="Replace the file inline instead of returning the output to stdout",
+    )
+    parser.add_argument(
+        "--ensure_ascii",
+        action="store_true",
+        help="Pass the ensure_ascii parameter to json.dumps()",
+    )
+    parser.add_argument(
+        "--indent",
+        type=int,
+        default=2,
+        help="Number of spaces for indentation (Default 2)",
+    )
+
+    args = parser.parse_args()
+
+    ensure_ascii = False
+    if args.ensure_ascii:
+        ensure_ascii = True
+    try:
+        result = from_file(args.filename)
+
+        if args.inline:
+            fd = open(args.filename, mode="w")
+            json.dump(result, fd, indent=args.indent, ensure_ascii=ensure_ascii)
+            fd.close()
+        else:
+            print(json.dumps(result, indent=args.indent, ensure_ascii=ensure_ascii))
+    except Exception as e:
+        print(f"Error: {str(e)}", file=sys.stderr)
+        sys.exit(1)
+
+
+if __name__ == "__main__":  # pragma: no cover
+    cli()

{json_repair-0.28.4.dist-info → json_repair-0.29.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.28.4
+Version: 0.29.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -150,6 +150,31 @@ Some rules of thumb to use:
 - `skip_json_loads` is faster only if you 100% know that the string is not a valid JSON
 - If you are having issues with escaping pass the string as **raw** string like: `r"string with escaping\""`
 
+### Use json_repair from CLI
+
+Install the library for command-line with:
+```
+pipx install json-repair
+```
+then run
+```
+$ json_repair -h
+
+usage: json_repair [-h] [-i] [--ensure_ascii] [--indent INDENT] filename
+
+Repair and parse JSON files.
+
+positional arguments:
+  filename         The JSON file to repair
+
+options:
+  -h, --help       show this help message and exit
+  -i, --inline     Replace the file inline instead of returning the output to stdout
+  --ensure_ascii   Pass the ensure_ascii parameter to json.dumps()
+  --indent INDENT  Number of spaces for indentation (Default 2)
+```
+to learn how to use it
+
 ## Adding to requirements
 **Please pin this library only on the major version!**
 

json_repair-0.29.0.dist-info/RECORD

@@ -0,0 +1,10 @@
+json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
+json_repair/__main__.py,sha256=EsJb-y89uZEvGQQg1GdIDWzfDwfOMvVekKEtdguQXCM,67
+json_repair/json_repair.py,sha256=hltJ3Qa4qFbUD3mVKkYvFWksnCcIZqx8zamKfBpjeNs,33538
+json_repair/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+json_repair-0.29.0.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
+json_repair-0.29.0.dist-info/METADATA,sha256=yh0EJo-I1u0R6X-Gq9ETz0WbgmuGIhzR7Icw9W4Kee0,9630
+json_repair-0.29.0.dist-info/WHEEL,sha256=uCRv0ZEik_232NlR4YDw4Pv3Ajt5bKvMH13NUU7hFuI,91
+json_repair-0.29.0.dist-info/entry_points.txt,sha256=SNfge3zPSP-ASqriYU9r3NAPaXdseYr7ciPMKdV2uSw,57
+json_repair-0.29.0.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
+json_repair-0.29.0.dist-info/RECORD,,

{json_repair-0.28.4.dist-info → json_repair-0.29.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (74.0.0)
+Generator: setuptools (74.1.1)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

json_repair-0.29.0.dist-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+json_repair = json_repair.__main__:cli

json_repair-0.28.4.dist-info/RECORD

@@ -1,8 +0,0 @@
-json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
-json_repair/json_repair.py,sha256=odtRiFJ-u8mbdw_3Djx4jADxGoBeQvot3536D6Y6K0c,32266
-json_repair/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-json_repair-0.28.4.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
-json_repair-0.28.4.dist-info/METADATA,sha256=2JB2TM0mrFC7OejTtgFrpyr2qQOw8xwX0KmLTSzephk,9019
-json_repair-0.28.4.dist-info/WHEEL,sha256=UvcQYKBHoFqaQd6LKyqHw9fxEolWLQnlzP0h_LgJAfI,91
-json_repair-0.28.4.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
-json_repair-0.28.4.dist-info/RECORD,,