json-repair 0.17.0__py3-none-any.whl → 0.17.2__py3-none-any.whl

json_repair/json_repair.py

@@ -227,6 +227,10 @@ class JSONParser:
             self.index += 1
             char = self.get_char_at()
 
+        if not char:
+            # This is an empty string
+            return ""
+
         # Ensuring we use the right delimiter
         if char == "'":
             lstring_delimiter = rstring_delimiter = "'"
@@ -298,16 +302,6 @@ class JSONParser:
             string_acc += char
             self.index += 1
             char = self.get_char_at()
-            # If the string contains an escaped character we should respect that or remove the escape
-            if self.get_char_at(-1) == "\\":
-                if char in [rstring_delimiter, "t", "n", "r", "b", "\\"]:
-                    string_acc += char
-                    self.index += 1
-                    char = self.get_char_at()
-                else:
-                    # Remove this character from the final output
-                    string_acc = string_acc[:-2] + string_acc[-1:]
-                    self.index -= 1
             # ChatGPT sometimes forget to quote stuff in html tags or markdown, so we do this whole thing here
             if char == rstring_delimiter:
                 # Special case here, in case of double quotes one after another
@@ -431,28 +425,40 @@ class JSONParser:
         return ""
 
     def get_char_at(self, count: int = 0) -> Union[str, bool]:
-        if self.json_fd:
-            self.json_fd.seek(self.index + count)
-            char = self.json_fd.read(1)
-            if char == "":
-                return False
-            return char
-        else:
-            # Why not use something simpler? Because we might be out of bounds and doing this check all the time is annoying
-            try:
-                return self.json_str[self.index + count]
-            except IndexError:
+        # Why not use something simpler? Because try/except in python is a faster alternative to an "if" statement that is often True
+        try:
+            return self.json_str[self.index + count]
+        except IndexError:
+            if self.json_fd:
+                self.json_fd.seek(self.index + count)
+                char = self.json_fd.read(1)
+                if char == "":
+                    return False
+                return char
+            else:
                 return False
 
     def skip_whitespaces_at(self) -> None:
         """
         This function quickly iterates on whitespaces, syntactic sugar to make the code more concise
         """
-
-        char = self.get_char_at()
-        while char and char.isspace():
-            self.index += 1
+        if self.json_fd:
             char = self.get_char_at()
+            while char and char.isspace():
+                self.index += 1
+                char = self.get_char_at()
+        else:
+            # If this is not a file stream, we do this monster here to make this function much much faster
+            try:
+                char = self.json_str[self.index]
+            except IndexError:
+                return
+            while char.isspace():
+                self.index += 1
+                try:
+                    char = self.json_str[self.index]
+                except IndexError:
+                    return
 
     def set_context(self, value: str) -> None:
         # If a value is provided update the context variable and save in stack
@@ -479,11 +485,17 @@ class JSONParser:
                 context = self.json_fd.read(self.logger["window"] * 2)
                 self.json_fd.seek(self.index)
             else:
-                context = self.json_str[
-                    self.index
-                    - self.logger["window"] : self.index
-                    + self.logger["window"]
-                ]
+                start = (
+                    self.index - self.logger["window"]
+                    if (self.index - self.logger["window"]) >= 0
+                    else 0
+                )
+                end = (
+                    self.index + self.logger["window"]
+                    if (self.index + self.logger["window"]) <= len(self.json_str)
+                    else len(self.json_str)
+                )
+                context = self.json_str[start:end]
             self.logger["log"].append(
                 {
                     "text": text,

{json_repair-0.17.0.dist-info → json_repair-0.17.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.17.0
+Version: 0.17.2
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -39,6 +39,12 @@ This simple package can be used to fix an invalid json string. To know all cases
 
 Inspired by https://github.com/josdejong/jsonrepair
 
+---
+# Offer me a beer
+If you find this library useful, you can help me by donating toward my monthly beer budget here: https://github.com/sponsors/mangiucugna
+
+---
+
 # Motivation
 Some LLMs are a bit iffy when it comes to returning well formed JSON data, sometimes they skip a parentheses and sometimes they add some words in it, because that's what an LLM does.
 Luckily, the mistakes LLMs make are simple enough to be fixed without destroying the content.
@@ -154,10 +160,6 @@ You will need owner access to this repository
 # Repair JSON in other programming languages
 - Typescript: https://github.com/josdejong/jsonrepair
 - Go: https://github.com/RealAlexandreAI/json-repair
----
-# Bonus Content
-If you need some good Custom Instructions (System Message) to improve your chatbot responses try https://gist.github.com/mangiucugna/7ec015c4266df11be8aa510be0110fe4
-
 ---
 ## Star History
 

json_repair-0.17.2.dist-info/RECORD

@@ -0,0 +1,7 @@
+json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
+json_repair/json_repair.py,sha256=NeAKskIWOewTpIDY3Z6wxYkDHjrNYHNuF4AFO-xBggI,23052
+json_repair-0.17.2.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
+json_repair-0.17.2.dist-info/METADATA,sha256=5YoWE0n-iKY2nbZ0D9OQ8QwfA6zZIYVWUkLsYQmxrqo,7333
+json_repair-0.17.2.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+json_repair-0.17.2.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
+json_repair-0.17.2.dist-info/RECORD,,

json_repair-0.17.0.dist-info/RECORD

@@ -1,7 +0,0 @@
-json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
-json_repair/json_repair.py,sha256=BeThFqLxt2-ln7SA3UPBI47VyxnV2MVDFEaolIOCiNU,22721
-json_repair-0.17.0.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
-json_repair-0.17.0.dist-info/METADATA,sha256=F2KOc46utpaJg3eqlw8oaWhisj3rDnHk-k-5miiWL3U,7355
-json_repair-0.17.0.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-json_repair-0.17.0.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
-json_repair-0.17.0.dist-info/RECORD,,