json-repair 0.19.1__py3-none-any.whl → 0.20.0__py3-none-any.whl

json_repair/json_repair.py

@@ -22,6 +22,7 @@ If something is wrong (a missing parantheses or quotes for example) it will use
 All supported use cases are in the unit tests
 """
 
+import os
 import json
 from typing import Any, Dict, List, Union, TextIO
 
@@ -31,7 +32,9 @@ class JSONParser:
         # The string to parse
         self.json_str = json_str
         # Alternatively, the file description with a json file in it
-        self.json_fd = json_fd
+        if json_fd:
+            # This is a trick we do to treat the file wrapper as an array
+            self.json_str = StringFileWrapper(json_fd)
         # Index is our iterator that will keep track of which character we are looking at right now
         self.index = 0
         # This is used in the object member parsing to manage the special cases of missing quotes in key or value
@@ -246,7 +249,8 @@ class JSONParser:
             rstring_delimiter = "”"
         elif char.isalpha():
             # This could be a <boolean> and not a string. Because (T)rue or (F)alse or (N)ull are valid
-            if char.lower() in ["t", "f", "n"]:
+            # But remember, object keys are only of type string
+            if char.lower() in ["t", "f", "n"] and self.get_context() != "object_key":
                 value = self.parse_boolean_or_null()
                 if value != "":
                     return value
@@ -263,7 +267,8 @@ class JSONParser:
                 self.index += 1
                 return self.parse_json()
             self.log(
-                "While parsing a string, we found no starting quote, ignoring", "info"
+                "While parsing a string, we found no starting quote. Will add the quote back",
+                "info",
             )
             missing_quotes = True
 
@@ -330,32 +335,50 @@ class JSONParser:
                     # Check if eventually there is a rstring delimiter, otherwise we bail
                     i = 1
                     next_c = self.get_char_at(i)
-                    while next_c and next_c != rstring_delimiter:
+                    check_comma_in_object_value = True
+                    while next_c and next_c not in [
+                        rstring_delimiter,
+                        lstring_delimiter,
+                    ]:
+                        # This is a bit of a weird workaround, essentially in object_value context we don't always break on commas
+                        # This is because the routine after will make sure to correct any bad guess and this solves a corner case
+                        if next_c.isalpha():
+                            check_comma_in_object_value = False
                         # If we are in an object context, let's check for the right delimiters
                         if (
-                            next_c == lstring_delimiter
-                            or ("object_key" in self.context and next_c == ":")
-                            or ("object_value" in self.context and next_c in ["}", ","])
+                            ("object_key" in self.context and next_c in [":", "}"])
+                            or ("object_value" in self.context and next_c == "}")
                             or ("array" in self.context and next_c in ["]", ","])
+                            or (
+                                check_comma_in_object_value
+                                and self.get_context() == "object_value"
+                                and next_c == ","
+                            )
                         ):
                             break
                         i += 1
                         next_c = self.get_char_at(i)
                     if next_c == rstring_delimiter:
-                        # But this might not be it! This could be just a missing comma
-                        # We need to check if we find a rstring_delimiter and a colon after
-                        i += 1
-                        next_c = self.get_char_at(i)
-                        while next_c and next_c != rstring_delimiter:
+                        if self.get_context() == "object_value":
+                            # But this might not be it! This could be just a missing comma
+                            # We found a delimiter and we need to check if this is a key
+                            # so find a rstring_delimiter and a colon after
                             i += 1
                             next_c = self.get_char_at(i)
-                        i += 1
-                        next_c = self.get_char_at(i)
-                        while next_c and next_c != ":":
-                            if next_c in [lstring_delimiter, rstring_delimiter, ","]:
-                                break
+                            while next_c and next_c != rstring_delimiter:
+                                i += 1
+                                next_c = self.get_char_at(i)
                             i += 1
                             next_c = self.get_char_at(i)
+                            while next_c and next_c != ":":
+                                if next_c in [
+                                    lstring_delimiter,
+                                    rstring_delimiter,
+                                    ",",
+                                ]:
+                                    break
+                                i += 1
+                                next_c = self.get_char_at(i)
                         # Only if we fail to find a ':' then we know this is misplaced quote
                         if next_c != ":":
                             self.log(
@@ -451,36 +474,22 @@ class JSONParser:
         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
+            return False
 
     def skip_whitespaces_at(self) -> None:
         """
         This function quickly iterates on whitespaces, syntactic sugar to make the code more concise
         """
-        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
-            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
@@ -502,22 +511,9 @@ class JSONParser:
     def log(self, text: str, level: str) -> None:
         if level == self.logger["log_level"]:
             context = ""
-            if self.json_fd:
-                self.json_fd.seek(self.index - self.logger["window"])
-                context = self.json_fd.read(self.logger["window"] * 2)
-                self.json_fd.seek(self.index)
-            else:
-                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]
+            start = max(self.index - self.logger["window"], 0)
+            end = min(self.index + self.logger["window"], len(self.json_str))
+            context = self.json_str[start:end]
             self.logger["log"].append(
                 {
                     "text": text,
@@ -593,3 +589,31 @@ def from_file(
     fd.close()
 
     return jsonobj
+
+
+class StringFileWrapper:
+    # This is a trick to simplify the code above, transform the filedescriptor handling into an array handling
+    def __init__(self, fd: TextIO) -> None:
+        self.fd = fd
+        self.length = None
+
+    def __getitem__(self, index: int) -> Any:
+        if isinstance(index, slice):
+            self.fd.seek(index.start)
+            value = self.fd.read(index.stop - index.start)
+            self.fd.seek(index.start)
+            return value
+        else:
+            self.fd.seek(index)
+            return self.fd.read(1)
+
+    def __len__(self) -> int:
+        if not self.length:
+            current_position = self.fd.tell()
+            self.fd.seek(0, os.SEEK_END)
+            self.length = self.fd.tell()
+            self.fd.seek(current_position)
+        return self.length
+
+    def __setitem__(self):
+        raise Exception("This is read-only!")

{json_repair-0.19.1.dist-info → json_repair-0.20.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.19.1
+Version: 0.20.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License

json_repair-0.20.0.dist-info/RECORD

@@ -0,0 +1,7 @@
+json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
+json_repair/json_repair.py,sha256=zYg4tIwZ4rdVkCQ5XVceNQaOz2MT50O7jHJbJ1EpKhk,25446
+json_repair-0.20.0.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
+json_repair-0.20.0.dist-info/METADATA,sha256=2WrsjORPx37e4CqdkdiARwB-VoP5EyjsZaS0hcVnVBo,7333
+json_repair-0.20.0.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+json_repair-0.20.0.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
+json_repair-0.20.0.dist-info/RECORD,,

json_repair-0.19.1.dist-info/RECORD

@@ -1,7 +0,0 @@
-json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
-json_repair/json_repair.py,sha256=5viwB7G8T9EyYRsJpCxZPZChmgDkaT1WQ7oxFUjm5Lg,24349
-json_repair-0.19.1.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
-json_repair-0.19.1.dist-info/METADATA,sha256=FlAtvarNR4sXKlNYr7VDedyScl7t8UgqyDyNZ2NIS10,7333
-json_repair-0.19.1.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-json_repair-0.19.1.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
-json_repair-0.19.1.dist-info/RECORD,,