change arxiv download attempt url order

* Update bridge_all.py

* Update bridge_chatgpt.py

* Update bridge_chatgpt.py

* Update bridge_all.py

* Update bridge_all.py

.github/workflows/build-with-latex-arm.yml

@@ -0,0 +1,51 @@
+# https://docs.github.com/en/actions/publishing-packages/publishing-docker-images#publishing-images-to-github-packages
+name: build-with-latex-arm
+
+on:
+  push:
+    branches:
+      - "master"
+
+env:
+  REGISTRY: ghcr.io
+  IMAGE_NAME: ${{ github.repository }}_with_latex_arm
+
+jobs:
+  build-and-push-image:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      packages: write
+
+    steps:
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Log in to the Container registry
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Extract metadata (tags, labels) for Docker
+        id: meta
+        uses: docker/metadata-action@v4
+        with:
+          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
+
+      - name: Build and push Docker image
+        uses: docker/build-push-action@v6
+        with:
+          context: .
+          push: true
+          platforms: linux/arm64
+          file: docs/GithubAction+NoLocal+Latex+Arm
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}

crazy_functions/Latex_Function.py

@@ -138,25 +138,43 @@ def arxiv_download(chatbot, history, txt, allow_cache=True):
     cached_translation_pdf = check_cached_translation_pdf(arxiv_id)
     if cached_translation_pdf and allow_cache: return cached_translation_pdf, arxiv_id
 
-    url_tar = url_.replace('/abs/', '/e-print/')
-    translation_dir = pj(ARXIV_CACHE_DIR, arxiv_id, 'e-print')
     extract_dst = pj(ARXIV_CACHE_DIR, arxiv_id, 'extract')
-    os.makedirs(translation_dir, exist_ok=True)
-
-    # <-------------- download arxiv source file ------------->
+    translation_dir = pj(ARXIV_CACHE_DIR, arxiv_id, 'e-print')
     dst = pj(translation_dir, arxiv_id + '.tar')
-    if os.path.exists(dst):
-        yield from update_ui_lastest_msg("调用缓存", chatbot=chatbot, history=history)  # 刷新界面
+    os.makedirs(translation_dir, exist_ok=True)
+    # <-------------- download arxiv source file ------------->
+
+    def fix_url_and_download():
+        # for url_tar in [url_.replace('/abs/', '/e-print/'), url_.replace('/abs/', '/src/')]:
+        for url_tar in [url_.replace('/abs/', '/src/'), url_.replace('/abs/', '/e-print/')]:
+            proxies = get_conf('proxies')
+            r = requests.get(url_tar, proxies=proxies)
+            if r.status_code == 200:
+                with open(dst, 'wb+') as f:
+                    f.write(r.content)
+                return True
+        return False
+
+    if os.path.exists(dst) and allow_cache:
+        yield from update_ui_lastest_msg(f"调用缓存 {arxiv_id}", chatbot=chatbot, history=history)  # 刷新界面
+        success = True
     else:
-        yield from update_ui_lastest_msg("开始下载", chatbot=chatbot, history=history)  # 刷新界面
-        proxies = get_conf('proxies')
-        r = requests.get(url_tar, proxies=proxies)
-        with open(dst, 'wb+') as f:
-            f.write(r.content)
+        yield from update_ui_lastest_msg(f"开始下载 {arxiv_id}", chatbot=chatbot, history=history)  # 刷新界面
+        success = fix_url_and_download()
+        yield from update_ui_lastest_msg(f"下载完成 {arxiv_id}", chatbot=chatbot, history=history)  # 刷新界面
+
+
+    if not success:
+        yield from update_ui_lastest_msg(f"下载失败 {arxiv_id}", chatbot=chatbot, history=history)
+        raise tarfile.ReadError(f"论文下载失败 {arxiv_id}")
+
     # <-------------- extract file ------------->
-    yield from update_ui_lastest_msg("下载完成", chatbot=chatbot, history=history)  # 刷新界面
     from toolbox import extract_archive
-    extract_archive(file_path=dst, dest_dir=extract_dst)
+    try:
+        extract_archive(file_path=dst, dest_dir=extract_dst)
+    except tarfile.ReadError:
+        os.remove(dst)
+        raise tarfile.ReadError(f"论文下载失败")
     return extract_dst, arxiv_id
 
 

crazy_functions/latex_fns/latex_toolbox.py

@@ -644,6 +644,213 @@ def run_in_subprocess(func):
 
 
 def _merge_pdfs(pdf1_path, pdf2_path, output_path):
+    try:
+        logger.info("Merging PDFs using _merge_pdfs_ng")
+        _merge_pdfs_ng(pdf1_path, pdf2_path, output_path)
+    except:
+        logger.info("Merging PDFs using _merge_pdfs_legacy")
+        _merge_pdfs_legacy(pdf1_path, pdf2_path, output_path)
+
+
+def _merge_pdfs_ng(pdf1_path, pdf2_path, output_path):
+    import PyPDF2  # PyPDF2这个库有严重的内存泄露问题，把它放到子进程中运行，从而方便内存的释放
+    from PyPDF2.generic import NameObject, TextStringObject, ArrayObject, FloatObject, NumberObject
+
+    Percent = 1
+    # raise RuntimeError('PyPDF2 has a serious memory leak problem, please use other tools to merge PDF files.')
+    # Open the first PDF file
+    with open(pdf1_path, "rb") as pdf1_file:
+        pdf1_reader = PyPDF2.PdfFileReader(pdf1_file)
+        # Open the second PDF file
+        with open(pdf2_path, "rb") as pdf2_file:
+            pdf2_reader = PyPDF2.PdfFileReader(pdf2_file)
+            # Create a new PDF file to store the merged pages
+            output_writer = PyPDF2.PdfFileWriter()
+            # Determine the number of pages in each PDF file
+            num_pages = max(pdf1_reader.numPages, pdf2_reader.numPages)
+            # Merge the pages from the two PDF files
+            for page_num in range(num_pages):
+                # Add the page from the first PDF file
+                if page_num < pdf1_reader.numPages:
+                    page1 = pdf1_reader.getPage(page_num)
+                else:
+                    page1 = PyPDF2.PageObject.createBlankPage(pdf1_reader)
+                # Add the page from the second PDF file
+                if page_num < pdf2_reader.numPages:
+                    page2 = pdf2_reader.getPage(page_num)
+                else:
+                    page2 = PyPDF2.PageObject.createBlankPage(pdf1_reader)
+                # Create a new empty page with double width
+                new_page = PyPDF2.PageObject.createBlankPage(
+                    width=int(
+                        int(page1.mediaBox.getWidth())
+                        + int(page2.mediaBox.getWidth()) * Percent
+                    ),
+                    height=max(page1.mediaBox.getHeight(), page2.mediaBox.getHeight()),
+                )
+                new_page.mergeTranslatedPage(page1, 0, 0)
+                new_page.mergeTranslatedPage(
+                    page2,
+                    int(
+                        int(page1.mediaBox.getWidth())
+                        - int(page2.mediaBox.getWidth()) * (1 - Percent)
+                    ),
+                    0,
+                )
+                if "/Annots" in new_page:
+                    annotations = new_page["/Annots"]
+                    for i, annot in enumerate(annotations):
+                        annot_obj = annot.get_object()
+
+                        # 检查注释类型是否是链接（/Link）
+                        if annot_obj.get("/Subtype") == "/Link":
+                            # 检查是否为内部链接跳转（/GoTo）或外部URI链接（/URI）
+                            action = annot_obj.get("/A")
+                            if action:
+
+                                if "/S" in action and action["/S"] == "/GoTo":
+                                    # 内部链接：跳转到文档中的某个页面
+                                    dest = action.get("/D")  # 目标页或目标位置
+                                    # if dest and annot.idnum in page2_annot_id:
+                                    if dest in pdf2_reader.named_destinations:
+                                        # 获取原始文件中跳转信息，包括跳转页面
+                                        destination = pdf2_reader.named_destinations[
+                                            dest
+                                        ]
+                                        page_number = (
+                                            pdf2_reader.get_destination_page_number(
+                                                destination
+                                            )
+                                        )
+                                        # 更新跳转信息，跳转到对应的页面和，指定坐标 (100, 150)，缩放比例为 100%
+                                        # “/D”:[10,'/XYZ',100,100,0]
+                                        if destination.dest_array[1] == "/XYZ":
+                                            annot_obj["/A"].update(
+                                                {
+                                                    NameObject("/D"): ArrayObject(
+                                                        [
+                                                            NumberObject(page_number),
+                                                            destination.dest_array[1],
+                                                            FloatObject(
+                                                                destination.dest_array[
+                                                                    2
+                                                                ]
+                                                                + int(
+                                                                    page1.mediaBox.getWidth()
+                                                                )
+                                                            ),
+                                                            destination.dest_array[3],
+                                                            destination.dest_array[4],
+                                                        ]
+                                                    )  # 确保键和值是 PdfObject
+                                                }
+                                            )
+                                        else:
+                                            annot_obj["/A"].update(
+                                                {
+                                                    NameObject("/D"): ArrayObject(
+                                                        [
+                                                            NumberObject(page_number),
+                                                            destination.dest_array[1],
+                                                        ]
+                                                    )  # 确保键和值是 PdfObject
+                                                }
+                                            )
+
+                                        rect = annot_obj.get("/Rect")
+                                        # 更新点击坐标
+                                        rect = ArrayObject(
+                                            [
+                                                FloatObject(
+                                                    rect[0]
+                                                    + int(page1.mediaBox.getWidth())
+                                                ),
+                                                rect[1],
+                                                FloatObject(
+                                                    rect[2]
+                                                    + int(page1.mediaBox.getWidth())
+                                                ),
+                                                rect[3],
+                                            ]
+                                        )
+                                        annot_obj.update(
+                                            {
+                                                NameObject(
+                                                    "/Rect"
+                                                ): rect  # 确保键和值是 PdfObject
+                                            }
+                                        )
+                                    # if dest and annot.idnum in page1_annot_id:
+                                    if dest in pdf1_reader.named_destinations:
+
+                                        # 获取原始文件中跳转信息，包括跳转页面
+                                        destination = pdf1_reader.named_destinations[
+                                            dest
+                                        ]
+                                        page_number = (
+                                            pdf1_reader.get_destination_page_number(
+                                                destination
+                                            )
+                                        )
+                                        # 更新跳转信息，跳转到对应的页面和，指定坐标 (100, 150)，缩放比例为 100%
+                                        # “/D”:[10,'/XYZ',100,100,0]
+                                        if destination.dest_array[1] == "/XYZ":
+                                            annot_obj["/A"].update(
+                                                {
+                                                    NameObject("/D"): ArrayObject(
+                                                        [
+                                                            NumberObject(page_number),
+                                                            destination.dest_array[1],
+                                                            FloatObject(
+                                                                destination.dest_array[
+                                                                    2
+                                                                ]
+                                                            ),
+                                                            destination.dest_array[3],
+                                                            destination.dest_array[4],
+                                                        ]
+                                                    )  # 确保键和值是 PdfObject
+                                                }
+                                            )
+                                        else:
+                                            annot_obj["/A"].update(
+                                                {
+                                                    NameObject("/D"): ArrayObject(
+                                                        [
+                                                            NumberObject(page_number),
+                                                            destination.dest_array[1],
+                                                        ]
+                                                    )  # 确保键和值是 PdfObject
+                                                }
+                                            )
+
+                                        rect = annot_obj.get("/Rect")
+                                        rect = ArrayObject(
+                                            [
+                                                FloatObject(rect[0]),
+                                                rect[1],
+                                                FloatObject(rect[2]),
+                                                rect[3],
+                                            ]
+                                        )
+                                        annot_obj.update(
+                                            {
+                                                NameObject(
+                                                    "/Rect"
+                                                ): rect  # 确保键和值是 PdfObject
+                                            }
+                                        )
+
+                                elif "/S" in action and action["/S"] == "/URI":
+                                    # 外部链接：跳转到某个URI
+                                    uri = action.get("/URI")
+                output_writer.addPage(new_page)
+            # Save the merged PDF file
+            with open(output_path, "wb") as output_file:
+                output_writer.write(output_file)
+
+
+def _merge_pdfs_legacy(pdf1_path, pdf2_path, output_path):
     import PyPDF2  # PyPDF2这个库有严重的内存泄露问题，把它放到子进程中运行，从而方便内存的释放
 
     Percent = 0.95

docs/GithubAction+NoLocal+Latex+Arm

@@ -0,0 +1,25 @@
+# 此Dockerfile适用于“无本地模型”的环境构建，如果需要使用chatglm等本地模型，请参考 docs/Dockerfile+ChatGLM
+# - 1 修改 `config.py`
+# - 2 构建 docker build -t gpt-academic-nolocal-latex -f docs/GithubAction+NoLocal+Latex .
+# - 3 运行 docker run -v /home/fuqingxu/arxiv_cache:/root/arxiv_cache --rm -it --net=host gpt-academic-nolocal-latex
+
+FROM menghuan1918/ubuntu_uv_ctex:latest
+ENV DEBIAN_FRONTEND=noninteractive
+SHELL ["/bin/bash", "-c"]
+WORKDIR /gpt
+COPY . .
+RUN /root/.cargo/bin/uv venv --seed \
+    && source .venv/bin/activate \
+    && /root/.cargo/bin/uv pip install openai numpy arxiv rich colorama Markdown pygments pymupdf python-docx pdfminer \
+    && /root/.cargo/bin/uv pip install -r requirements.txt \
+    && /root/.cargo/bin/uv clean
+
+# 对齐python3
+RUN rm -f /usr/bin/python3 && ln -s /gpt/.venv/bin/python /usr/bin/python3
+RUN rm -f /usr/bin/python && ln -s /gpt/.venv/bin/python /usr/bin/python
+
+# 可选步骤，用于预热模块
+RUN python3 -c 'from check_proxy import warm_up_modules; warm_up_modules()'
+
+# 启动
+CMD ["python3", "-u", "main.py"]

request_llms/bridge_all.py

@@ -256,6 +256,8 @@ model_info = {
         "max_token": 128000,
         "tokenizer": tokenizer_gpt4,
         "token_cnt": get_token_num_gpt4,
+        "openai_disable_system_prompt": True,
+        "openai_disable_stream": True,
     },
     "o1-mini": {
         "fn_with_ui": chatgpt_ui,
@@ -264,6 +266,8 @@ model_info = {
         "max_token": 128000,
         "tokenizer": tokenizer_gpt4,
         "token_cnt": get_token_num_gpt4,
+        "openai_disable_system_prompt": True,
+        "openai_disable_stream": True,
     },
 
     "gpt-4-turbo": {
@@ -1281,4 +1285,3 @@ def predict(inputs:str, llm_kwargs:dict, plugin_kwargs:dict, chatbot,
 
     # 更新一下llm_kwargs的参数，否则会出现参数不匹配的问题
     yield from method(inputs, llm_kwargs, plugin_kwargs, chatbot, history, system_prompt, stream, additional_fn)
-

request_llms/bridge_chatgpt.py

@@ -202,10 +202,13 @@ def predict_no_ui_long_connection(inputs:str, llm_kwargs:dict, history:list=[],
                     if (time.time()-observe_window[1]) > watch_dog_patience:
                         raise RuntimeError("用户取消了程序。")
         else: raise RuntimeError("意外Json结构："+delta)
-    if json_data and json_data['finish_reason'] == 'content_filter':
-        raise RuntimeError("由于提问含不合规内容被Azure过滤。")
-    if json_data and json_data['finish_reason'] == 'length':
+
+    finish_reason = json_data.get('finish_reason', None) if json_data else None
+    if finish_reason == 'content_filter':
+        raise RuntimeError("由于提问含不合规内容被过滤。")
+    if finish_reason == 'length':
         raise ConnectionAbortedError("正常结束，但显示Token不足，导致输出不完整，请削减单次输入的文本量。")
+
     return result
 
 
@@ -536,4 +539,3 @@ def generate_payload(inputs:str, llm_kwargs:dict, history:list, system_prompt:st
 
     return headers,payload
 
-