utils/json_extractor.py

import json
from typing import Any, List

from logging_config import get_logger

logger = get_logger("utils.json_extractor")


def _find_balanced_closing_index(text: str, start_index: int) -> int:
    start_char = text[start_index]
    end_char = "}" if start_char == "{" else "]"

    depth = 0
    in_double_quote = False
    in_single_quote = False
    in_backtick = False
    in_line_comment = False
    in_block_comment = False
    is_escaped = False

    length = len(text)
    i = start_index

    while i < length:
        char = text[i]
        next_char = text[i + 1] if i + 1 < length else ""

        if is_escaped:
            is_escaped = False
            i += 1
            continue

        if char == "\\" and not in_line_comment and not in_block_comment:
            is_escaped = True
            i += 1
            continue

        if in_line_comment:
            if char == "\n":
                in_line_comment = False
            i += 1
            continue

        if in_block_comment:
            if char == "*" and next_char == "/":
                in_block_comment = False
                i += 2
                continue
            i += 1
            continue

        if not in_double_quote and not in_single_quote and not in_backtick:
            if char == "/" and next_char == "/":
                in_line_comment = True
                i += 2
                continue
            if char == "/" and next_char == "*":
                in_block_comment = True
                i += 2
                continue

        if in_double_quote:
            if char == '"':
                in_double_quote = False
            i += 1
            continue

        if in_single_quote:
            if char == "'":
                in_single_quote = False
            i += 1
            continue

        if in_backtick:
            if char == "`":
                in_backtick = False
            i += 1
            continue

        if char == '"':
            in_double_quote = True
            i += 1
            continue
        if char == "'":
            in_single_quote = True
            i += 1
            continue
        if char == "`":
            in_backtick = True
            i += 1
            continue

        if char == start_char:
            depth += 1
        elif char == end_char:
            depth -= 1
            if depth == 0:
                return i

        i += 1

    return -1


def extract_json_from_content(content: str) -> List[Any]:
    logger.debug("Starting JSON extraction from content of length %d", len(content))

    if not content or not isinstance(content, str):
        logger.warning("Received empty or non-string content for JSON extraction")
        return []

    found_blocks: List[Any] = []
    cursor = 0
    length = len(content)

    while cursor < length:
        if content[cursor] not in {"{", "["}:
            cursor += 1
            continue

        end_index = _find_balanced_closing_index(content, cursor)

        if end_index == -1:
            logger.debug("No balanced closing bracket found at cursor=%d", cursor)
            cursor += 1
            continue

        raw_candidate = content[cursor: end_index + 1]
        try:
            parsed = json.loads(raw_candidate)
            logger.debug("Successfully parsed JSON block at cursor=%d", cursor)
            found_blocks.append(parsed)
            cursor = end_index + 1
        except json.JSONDecodeError:
            cursor += 1

    logger.info("JSON extraction complete: found %d block(s)", len(found_blocks))
    return found_blocks