mseep-patche 1.0.1__py3-none-any.whl

Patche/utils/parse.py

@@ -0,0 +1,269 @@
+import re
+from typing import List, Optional
+
+from whatthepatch_pydantic import parse_patch as wtp_parse_patch
+from whatthepatch_pydantic.model import Diff as WTPDiff
+
+from Patche.config import settings
+from Patche.model import Change, Diff, Hunk, Patch
+from Patche.utils.header import CHANGE_LINE, HEADER_OLD, HUNK_START, parse_header
+
+git_diffcmd_header = re.compile("^diff --git a/(.+) b/(.+)$")
+unified_diff_header = re.compile("^---\s{1}")
+spliter_line = re.compile("^---$")
+
+
+def changes_to_hunks(changes: list[Change]) -> list[Hunk]:
+    """
+    Convert a list of changes to a list of hunks
+
+    Args:
+        changes (list[Change]): A list of changes
+
+    Returns:
+        list[Hunk]: A list of hunks
+    """
+
+    # 首先统计 Hunk 数
+    hunk_indexes = []
+    for change in changes:
+        if change.hunk not in hunk_indexes:
+            hunk_indexes.append(change.hunk)
+
+    # 将changes按照hunk分组，注意同一个 hunk 中的 change 要进行分类，前三行要放入前置上下文，中间的要放入中间上下文，后三行要放入后置上下文
+    hunk_list: list[Hunk] = []
+    for hunk_index in hunk_indexes:
+        hunk_changes = [change for change in changes if change.hunk == hunk_index]
+
+        # 这里遍历的顺序已经是正确的顺序
+        hunk_context = []
+        hunk_middle = []
+        hunk_post = []
+        # 首先正向遍历，获取前置上下文
+        for change in hunk_changes:
+            if change.old is not None and change.new is not None:
+                hunk_context.append(change)
+            else:
+                break
+
+        # 然后反向遍历，获取后置上下文
+        for change in reversed(hunk_changes):
+            if change.old is not None and change.new is not None:
+                hunk_post.append(change)
+            else:
+                break
+
+        assert len(hunk_context) <= settings.max_diff_lines
+        assert len(hunk_post) <= settings.max_diff_lines
+
+        # 最后获取中间代码
+        for change in hunk_changes:
+            if change not in hunk_context and change not in hunk_post:
+                hunk_middle.append(change)
+
+        # 注意把后置上下文反转回来
+        hunk_post = list(reversed(hunk_post))
+
+        hunk_list.append(
+            Hunk(
+                index=hunk_index,
+                context=hunk_context,
+                middle=hunk_middle,
+                post=hunk_post,
+                all_=hunk_changes,
+            )
+        )
+
+    return hunk_list
+
+
+def wtp_diff_to_diff(wtp_diff: WTPDiff) -> Diff:
+    """
+    Convert a whatthepatch Diff object to a patche Diff object
+
+    Args:
+        wtp_diff (WTPDiff): A whatthepatch Diff object
+
+    Returns:
+        Diff: A patche Diff object
+    """
+
+    return Diff(
+        header=wtp_diff.header,
+        changes=wtp_diff.changes,
+        text=wtp_diff.text,
+        hunks=changes_to_hunks(wtp_diff.changes),
+    )
+
+
+def parse_unified_diff(text: str) -> Optional[List[Diff]]:
+    """解析 unified diff 格式的补丁"""
+    lines = iter(text.splitlines())
+    diffs: List[Diff] = []
+
+    while True:
+        try:
+            header = parse_header(lines)
+            if not header:
+                break
+
+            changes: List[Change] = []
+            hunk_index = 0
+
+            for line in lines:
+                # 检查是否是新的 diff 块开始
+                if HEADER_OLD.match(line):
+                    lines = iter([line] + list(lines))
+                    break
+
+                # 解析 hunk 头
+                hunk_match = HUNK_START.match(line)
+                if hunk_match:
+                    old_start = int(hunk_match.group(1))
+                    old_count = int(hunk_match.group(2) or "1")
+                    new_start = int(hunk_match.group(3))
+                    new_count = int(hunk_match.group(4) or "1")
+
+                    old_current = old_start
+                    new_current = new_start
+                    hunk_index += 1
+                    continue
+
+                # 解析变更行
+                change_match = CHANGE_LINE.match(line)
+                if change_match:
+                    change_type = change_match.group(1)
+                    content = change_match.group(2)
+
+                    if change_type == " ":
+                        # 上下文行 / 中间行
+                        changes.append(
+                            Change(
+                                old=old_current,
+                                new=new_current,
+                                line=content,
+                                hunk=hunk_index,
+                            )
+                        )
+                        old_current += 1
+                        new_current += 1
+                    elif change_type == "-":
+                        # 删除行
+                        changes.append(
+                            Change(
+                                old=old_current, new=None, line=content, hunk=hunk_index
+                            )
+                        )
+                        old_current += 1
+                    elif change_type == "+":
+                        # 新增行
+                        changes.append(
+                            Change(
+                                old=None, new=new_current, line=content, hunk=hunk_index
+                            )
+                        )
+                        new_current += 1
+
+            if header and changes:
+                diffs.append(
+                    Diff(
+                        header=header,
+                        changes=changes,
+                        text=text,
+                        hunks=changes_to_hunks(changes),
+                    )
+                )
+
+        except StopIteration:
+            break
+
+    return diffs if diffs else None
+
+
+def parse_patch(text: str) -> Patch:
+    """
+    Parse a patch file
+    Diiference between this and whatthepatch.parse_patch is that this function also
+    returns the sha, author, date and message of the commit
+    """
+
+    lines = text.splitlines()
+
+    idx = 0
+    for i, line in enumerate(lines):
+        # 这里考虑 git log 格式和 git format-patch 格式
+        if (
+            git_diffcmd_header.match(line)
+            or spliter_line.match(line)
+            or unified_diff_header.match(line)
+        ):
+            idx = i
+            break
+
+    else:
+        # raise ValueError(
+        #     "No diff --git line found, check if the input is a valid patch"
+        # )
+        idx = len(lines) + 1
+
+    git_message_lines: list[str] = []
+    if idx == 0:
+        return Patch(
+            sha=None,
+            author=None,
+            date=None,
+            subject=None,
+            message=None,
+            # diff=[wtp_diff_to_diff(diff) for diff in wtp_parse_patch(text)],
+            diff=parse_unified_diff(text),
+        )
+    else:
+        git_message_lines = lines[:idx]
+
+    message = "\n".join(git_message_lines)
+
+    sha_line = git_message_lines.pop(0)
+    if sha_line.startswith("From ") or sha_line.startswith("commit "):
+        sha = sha_line.split(" ")[1]
+    else:
+        sha = None
+
+    author_line = git_message_lines.pop(0)
+    if author_line.startswith("Author: ") or author_line.startswith("From:"):
+        author = " ".join(author_line.split(" ")[1:])
+    else:
+        author = None
+
+    date_line = git_message_lines.pop(0)
+    if date_line.startswith("Date: "):
+        date_str = date_line.split("Date: ")[1]
+        # 解析 Thu, 7 Mar 2024 15:41:57 +0800 或 Tue Feb 2 16:07:37 2021 +0100
+        # if "," in date_str:
+        #     date_fromat = "%a, %d %b %Y %H:%M:%S %z"
+        # else:
+        #     date_fromat = "%a %b %d %H:%M:%S %Y %z"
+
+        # date = datetime.datetime.strptime(date_str.strip(), date_fromat)
+        date = date_str.strip()
+    else:
+        date = None
+
+    # 如果接下来的一行以 Subject 开头，则直接解析出 subject
+    if git_message_lines[0].startswith("Subject: "):
+        subject = git_message_lines.pop(0).split("Subject: ")[1]
+    else:
+        # 否则找到剩余的行里第一个非换行/非空行作为 subject
+        subject = None
+        for line in git_message_lines:
+            if line.strip() != "":
+                subject = line
+                break
+
+    return Patch(
+        sha=sha.strip() if sha else None,
+        author=author.strip() if author else None,
+        date=date,
+        subject=subject.strip() if subject else None,
+        message=message,
+        diff=[wtp_diff_to_diff(diff) for diff in wtp_parse_patch(text)],
+    )

Patche/utils/resolve.py

@@ -0,0 +1,168 @@
+from Patche.app import logger
+from Patche.config import settings
+from Patche.model import ApplyResult, Hunk, Line
+from Patche.utils.common import find_list_positions
+
+
+def apply_change(
+    hunk_list: list[Hunk],
+    target: list[Line],
+    reverse: bool = False,
+    flag_hunk_list: list[int] = None,
+    fuzz: int = 2,  # set fuzz=2 according GNU Patch
+) -> ApplyResult:
+    """Apply a diff to a target string."""
+
+    flag_hunk_list = [] if flag_hunk_list is None else flag_hunk_list
+
+    if fuzz > settings.max_diff_lines or fuzz < 0:
+        raise Exception(f"fuzz value should be less than {settings.max_diff_lines}")
+
+    if reverse:
+        for hunk in hunk_list:
+            for change in hunk.context + hunk.middle + hunk.post:
+                change.old, change.new = change.new, change.old
+
+    # 然后对每个hunk进行处理，添加偏移
+    failed_hunk_list: list[Hunk] = []
+    last_pos = None
+
+    last_offset = 0
+    line_count_diff = 0
+    for hunk in hunk_list:
+
+        current_hunk_fuzz = 0
+
+        while current_hunk_fuzz <= fuzz:
+
+            # hunk.context = hunk.context[1:]
+            # hunk.post = hunk.post[: fuzz - current_hunk_fuzz]
+
+            logger.debug(
+                f"current_fuzz: {current_hunk_fuzz} len(hunk.context): {len(hunk.context)} len(hunk.post): {len(hunk.post)}"
+            )
+
+            changes_to_search = hunk.context + hunk.middle + hunk.post
+            pos_list = find_list_positions(
+                [line.content for line in target],
+                [change.line for change in changes_to_search if change.old is not None],
+            )
+
+            if len(pos_list) != 0:
+                break
+
+            current_hunk_fuzz += 1
+
+            if current_hunk_fuzz <= fuzz:
+                hunk.context = hunk.context[1:]
+                hunk.post = hunk.post[: 3 - current_hunk_fuzz]
+
+        # 初始位置是 context 的第一个
+        # 注意，前几个有可能是空
+        pos_origin = None
+        for change in changes_to_search:
+            if change.old is not None:
+                pos_origin = change.old
+                break
+
+        # TODO: 这里不太对，要想一下怎么处理，不应该是加入 failed hunk list
+        # 仅在 -F 3 且只有添加行 的情况下出现（指与 GNU patch 行为不一致）
+        # 也可以看一下这样的情况有多少
+        if current_hunk_fuzz == fuzz and pos_origin is not None:
+            # failed_hunk_list.append(hunk)
+            # logger.debug(f"Could not determine pos_origin")
+            # logger.warning(f"Apply failed with hunk {hunk.index}")
+            # continue
+            for change in changes_to_search:
+                if change.new is not None:
+                    pos_origin = change.new
+                    break
+
+            # 使用上一次偏移加上行数变化差值
+            min_offset = last_offset
+        else:
+            if len(pos_list) == 0:
+                failed_hunk_list.append(hunk)
+                logger.debug(f"Could not determine proper position")
+                logger.warning(f"Apply failed with hunk {hunk.index}")
+                continue
+
+            offset_list = [
+                pos + 1 - pos_origin for pos in pos_list
+            ]  # 确认这里是否需要 1？
+
+            # 计算最小 offset
+            min_offset = None
+            for offset in offset_list:
+                if min_offset is None or abs(offset) < abs(min_offset):
+                    min_offset = offset
+
+            if reverse:
+                min_offset += line_count_diff
+                pos_origin -= line_count_diff
+
+        last_offset = min_offset
+
+        # 更新行数变化差值
+        hunk_add_count = sum(
+            1 for c in changes_to_search if c.old is None and c.new is not None
+        )
+        hunk_del_count = sum(
+            1 for c in changes_to_search if c.new is None and c.old is not None
+        )
+        line_count_diff += hunk_del_count - hunk_add_count
+
+        logger.info(
+            f"Apply hunk {hunk.index} with offset {min_offset} fuzz {current_hunk_fuzz} line_diff {line_count_diff}"
+        )
+
+        # 直接按照 pos 进行替换
+        # 选择 offset 最小的 pos
+        pos_new = pos_origin + min_offset - 1
+
+        # 处理 pos_new 小于 last_pos 的情况
+        logger.debug(f"pos_origin: {pos_origin}, last_pos: {last_pos}")
+        if last_pos is None:
+            last_pos = pos_new
+        elif pos_new < last_pos:
+            # 特别主要 pos_new 小于 last_pos 的情况
+            logger.warning(f"Apply failed with hunk {hunk.index}")
+            logger.error(f"pos: {pos_new} is greater than last_pos: {last_pos}")
+            failed_hunk_list.append(hunk)
+            continue
+        else:
+            last_pos = pos_new
+
+        old_lines = [
+            change.line
+            for change in hunk.context + hunk.middle + hunk.post
+            if change.old is not None
+        ]
+        new_lines = [
+            change.line
+            for change in hunk.context + hunk.middle + hunk.post
+            if change.new is not None
+        ]
+
+        # 检查 pos_new 位置的行是否和 old_lines 一致
+        for i in range(len(old_lines)):
+            if target[pos_new + i].content != old_lines[i]:
+                raise Exception(
+                    f'line {pos_new + i}, "{target[pos_new + i].content}" does not match "{old_lines[i]}"'
+                )
+
+        # 以切片的方式进行替换
+        target = (
+            target[:pos_new]
+            + [
+                Line(index=pos_new + i, content=new_lines[i])
+                for i in range(len(new_lines))
+            ]
+            + target[pos_new + len(old_lines) :]
+        )
+
+    return ApplyResult(
+        new_line_list=target,
+        conflict_hunk_num_list=[],
+        failed_hunk_list=failed_hunk_list,
+    )

mseep_patche-1.0.1.dist-info/METADATA

@@ -0,0 +1,14 @@
+Metadata-Version: 2.1
+Name: mseep-patche
+Version: 1.0.1
+Summary: Modern Patch in Python
+Author-Email: mseep <support@skydeck.ai>
+License: MIT
+Requires-Python: >=3.10
+Requires-Dist: typer[all]>=0.9.0
+Requires-Dist: pydantic>=2.5.3
+Requires-Dist: pydantic-settings>=2.2.1
+Requires-Dist: whatthepatch-pydantic==1.0.6a3
+Description-Content-Type: text/plain
+
+Package managed by MseeP.ai

mseep_patche-1.0.1.dist-info/RECORD

@@ -0,0 +1,24 @@
+Patche/__init__.py,sha256=PVoMA8KWQKPFthAyEIn4oNZATeGNTGXVuqLfqoOjfmE,34
+Patche/__main__.py,sha256=QXfE3TKpFZsct7wDbpv5g3tQGmrUkBMjL_nXg_wJvEM,28
+Patche/__version__.py,sha256=wtfcRFm73_vwwNMAiQol_VGC7QetgMQWRsGYiTuqF0c,185
+Patche/app.py,sha256=I5AjgCmHmIE6jPaYYdbRwBL4dv5BK3nB4fKFDgI3u_w,838
+Patche/commands/apply.py,sha256=POcKV3oxtqD6jbkrujKycOnG9ruk1Wj6QadXlNQCb4U,3509
+Patche/commands/help.py,sha256=nEGX9GHT2CJHchLh02ZJ4RxPwWf84j6j4YpSFwS6cLo,214
+Patche/commands/show.py,sha256=Io-9vruhgrPf5GalJTfEm6KNseQ2LY6Eu0MiDLtcSFs,1105
+Patche/config.py,sha256=6k9RZ8skRTQ8hPYI1I63PkupsLkZSCmVxIhJZdwRVNU,499
+Patche/mcp/__init__.py,sha256=Y6bkduOn9Ljy76x8bOFGrcjx4RFB8zrY_FCVC7grfEY,815
+Patche/mcp/__main__.py,sha256=3xkjNDGmST-F8nTIpnE0OVBBBGL3sjToNbc-QRebi9k,36
+Patche/mcp/model.py,sha256=p_uTVFVB9PubhFJaUbqNeOlJr63WvTK70lmLDx7uLxM,472
+Patche/mcp/prompts.py,sha256=VFc0h6ayMmWvHg-NKTJgoCFS7KzSA6ttzy5JK8TT7nM,7726
+Patche/mcp/server.py,sha256=5fhkLIAgC5jf6qGgV0dh9lqEd13VvycUaBTdPjXYt_Y,3450
+Patche/mcp/tools.py,sha256=u7gNMd9NEXGpI20atx3pIEjeZLMCo6bUvtrSs_eoFy8,4886
+Patche/model.py,sha256=Dm-LneXe3idcxNRbAhZ01Z6hqw-HDioHeJ07h2aPdvM,1842
+Patche/utils/common.py,sha256=_4o3z2Dq-9qG2ahX3UmIfShTO1mesZ3DXGu0EIeS6BE,2020
+Patche/utils/header.py,sha256=t4wlitYDaZRq2zGlP0FqIP2ihz83eg_7jJ9VDuXnIYY,1075
+Patche/utils/parse.py,sha256=RVWxCAbweEDN7y6RQ44R-rRw-GDLFvRamMFxW9hOveA,8573
+Patche/utils/resolve.py,sha256=4bpUYtSfDK5dwBldw4essjirI4zbMOxutiMBnhPubkw,5854
+mseep_patche-1.0.1.dist-info/METADATA,sha256=GA-Wovcm8kAphOFhOuZs-zFpStPivRMS53Z_Jyyk5V4,381
+mseep_patche-1.0.1.dist-info/WHEEL,sha256=9P2ygRxDrTJz3gsagc0Z96ukrxjr-LFBGOgv3AuKlCA,90
+mseep_patche-1.0.1.dist-info/entry_points.txt,sha256=OhS7Z9mx7By8K1v0_XTqzwGs7v5BQEocPpu20F0WrOE,100
+mseep_patche-1.0.1.dist-info/licenses/LICENSE,sha256=2I7sSSPi_J8S7pOQIvPdaRQubMmIpsDI6qNblsWv5ek,1066
+mseep_patche-1.0.1.dist-info/RECORD,,

mseep_patche-1.0.1.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: pdm-backend (2.4.5)
+Root-Is-Purelib: true
+Tag: py3-none-any

mseep_patche-1.0.1.dist-info/entry_points.txt

@@ -0,0 +1,6 @@
+[console_scripts]
+patche = Patche.__init__:app
+patche-mcp = Patche.mcp.__init__:app
+
+[gui_scripts]
+

mseep_patche-1.0.1.dist-info/licenses/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 jingfelix
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.