Correções de execução do markapi. by gitnnolabs · Pull Request #39 · scieloorg/markapi

gitnnolabs · 2025-11-30T14:17:43Z

O que esse PR faz?

Esse PR realiza ajustes de instalação e execução do markapi.

Onde a revisão poderia começar?

Sugiro que para realizar a revisão seja realizado os seguintes passos:

Verificar os arquivos alterados nesse PR
Baixar o projeto no git
Mudar para a branch fix-markupapp
Executar os passos descritos no link: https://github.com/scieloorg/markapi/wiki/Guia-de-Instala%C3%A7%C3%A3o-do-MarkAPI
Executar a carga de um dos documentos incluídos na nova pasta fixtures para teste.
Em alguns documentos é possível no log do celery works verificar alguns erros, esse bug esta notificados no issue: Erro na marcação de arquivos DOCx #38

Como este poderia ser testado manualmente?

Para realizar os testes de marcação de um documento manualmente é necessário seguir os seguintes passos: https://github.com/scieloorg/markapi/wiki/Converter-DOCX-para-XML, considere o documento e14790.docx na fixtures que está sem erro de execução.

Algum cenário de contexto que queira dar?

Removi a duplicidade de containers e tornei o modelo llama para marcação obrigatório.

Screenshots

Segue algumas telas de documentos marcados:

Quais são tickets relevantes?

Não existe tíquete para essa atividade.

Referências

N/A

…ação do markup_doc e atribuições realizado nas tarefas.

…á necessário um modelo llama.

Copilot

Pull request overview

This PR addresses installation and execution issues for markapi by consolidating llama dependencies into base requirements and making the llama model mandatory for document markup operations. The changes remove duplicate Docker configurations and add test fixtures for validation.

Key changes:

Moved llama-related dependencies from optional to required base dependencies
Changed LLAMA_ENABLED default to True, making llama model mandatory
Renamed database field from status to estatus in Reference model
Removed duplicate Docker configurations (llama.local.yml and Dockerfile.llama)
Added test fixture DOCX files for document processing validation

Reviewed changes

Copilot reviewed 7 out of 10 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
requirements/base.txt	Consolidated llama dependencies (huggingface-hub, llama-cpp-python) into base requirements
requirements/extra-llama.txt	Removed file as dependencies moved to base
config/settings/base.py	Changed LLAMA_ENABLED default from False to True
reference/models.py	Renamed field from `status` to `estatus`
reference/wagtail_hooks.py	Updated field reference to match rename
llama.local.yml	Removed duplicate Docker Compose configuration
compose/local/django/Dockerfile.llama	Removed specialized Dockerfile
fixtures/e740.docx	Added test fixture (binary file)
fixtures/e14790.docx	Added test fixture (binary file)

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-12-08T14:40:55Z

reference/wagtail_hooks.py

                new_reference = Reference.objects.create(
                    mixed_citation=linea,
-                    status=1,  # Estatus predeterminado
+                    estatus=1,  # Estatus predeterminado


This field rename from status=1 to estatus=1 is a breaking change that requires a database migration. Ensure that a migration file has been created and tested to handle this field rename, otherwise existing database records will be inaccessible.

Copilot · 2025-12-08T14:40:55Z

config/settings/base.py


 # LLAMA
-LLAMA_ENABLED = env.bool("LLAMA_ENABLED", default=False)
+LLAMA_ENABLED = env.bool("LLAMA_ENABLED", default=True)


Changing LLAMA_ENABLED default from False to True makes the llama model mandatory for all deployments. This is a significant behavioral change that:

Requires the llama model file to be present in all environments

May impact existing deployments that don't have llama configured

Should be clearly documented in deployment/upgrade instructions

Consider whether this should be an opt-in change rather than changing the default, or ensure proper deployment documentation is updated.

Suggested change

LLAMA_ENABLED = env.bool("LLAMA_ENABLED", default=True)

LLAMA_ENABLED = env.bool("LLAMA_ENABLED", default=False)

Copilot · 2025-12-08T14:40:56Z

reference/models.py

    mixed_citation = models.TextField(_("Mixed Citation"), null=False, blank=True)

-    status = models.IntegerField(
+    estatus = models.IntegerField(


The field name change from status to estatus appears inconsistent with the model name ReferenceStatus and the translation _("Reference status"). This creates confusion as the field name suggests Spanish while the related choices class and translation key use English. Consider either:

Keeping the field name as status to match the choices class name

Or changing both to use consistent Spanish naming (e.g., ReferenceEstatus choices class)

Suggested change

estatus = models.IntegerField(

status = models.IntegerField(

Copilot

Pull request overview

Copilot reviewed 7 out of 10 changed files in this pull request and generated 2 comments.

Comments suppressed due to low confidence (1)

reference/models.py:29

Renaming Reference.status to Reference.estatus requires a schema migration (ideally a RenameField migration to preserve data). Right now existing migrations and code still reference status (e.g., reference/migrations/0001_initial.py, reference/api/v1/views.py, reference/data_utils.py), which will break at runtime after migrating. Add/update migrations and update all remaining status references to estatus (or keep the field name as status if the rename is not intended).

class Reference(CommonControlField, ClusterableModel):
    mixed_citation = models.TextField(_("Mixed Citation"), null=False, blank=True)

    estatus = models.IntegerField(
        _("Reference status"),
        choices=ReferenceStatus.choices,
        blank=True,
        default=ReferenceStatus.NO_REFERENCE
    )

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-24T12:25:09Z

requirements/base.txt

+# huggingface-hub
+# ------------------------------------------------------------------------------
+huggingface_hub==0.26.1  # https://pypi.org/project/huggingface-hub/
+
+# Python Bindings for llama.cpp
+# ------------------------------------------------------------------------------
+llama-cpp-python==0.3.14  # https://llama-cpp-python.readthedocs.io/en/latest/


huggingface_hub is not the pip package name (PyPI uses huggingface-hub). As written, installs will fail with “No matching distribution found”. Rename the requirement entry back to huggingface-hub==0.26.1 (the import module is huggingface_hub, but the distribution name uses a hyphen).

Copilot · 2026-03-24T12:25:09Z

reference/wagtail_hooks.py

                new_reference = Reference.objects.create(
                    mixed_citation=linea,
-                    status=1,  # Estatus predeterminado
+                    estatus=1,  # Estatus predeterminado
                    creator=self.request.user,  # Usuario asociado
                )


This creates references with a hard-coded estatus=1. Since ReferenceStatus exists and is used elsewhere, prefer using the enum value (e.g., ReferenceStatus.CREATING) to avoid magic numbers and keep behavior stable if the enum changes.

Copilot

Pull request overview

Copilot reviewed 7 out of 10 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (1)

reference/models.py:29

O campo do modelo foi renomeado de status para estatus, mas as migrações existentes (ex.: reference/migrations/0001_initial.py) ainda criam a coluna status e há usos do atributo status em outras partes do app (ex.: reference/data_utils.py, reference/api/v1/views.py). Isso vai quebrar em runtime (AttributeError) e/ou na migração do schema. Sugestão: manter o nome status no model ou fazer um RenameField via migration e atualizar todos os acessos/serializers/queries para o novo nome de forma consistente.

    estatus = models.IntegerField(
        _("Reference status"),
        choices=ReferenceStatus.choices,
        blank=True,
        default=ReferenceStatus.NO_REFERENCE
    )

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-24T12:48:16Z

reference/wagtail_hooks.py

                new_reference = Reference.objects.create(
                    mixed_citation=linea,
-                    status=1,  # Estatus predeterminado
+                    estatus=1,  # Estatus predeterminado


Aqui foi trocado para estatus=1, mas o restante do app ainda usa status (e as migrações criam status). Isso vai falhar ao criar o objeto via Wagtail. Sugestão: usar status=ReferenceStatus.CREATING (ou equivalente) ou alinhar a renomeação do campo com migration + atualização completa do código.

Suggested change

estatus=1, # Estatus predeterminado

status=1, # Estatus predeterminado

gitnnolabs added 6 commits November 30, 2025 10:48

Torna a execução das tarefas de marcação obrigatório o uso do Llama.

77ccda1

Corrigi o nome do atributo status para estatusconforme a migr…

bbee01a

…ação do markup_doc e atribuições realizado nas tarefas.

Corrigi o atributo status para estatus na view das referencias.

c161e33

Assegura que sempre iremos ter o llama para marcação dos arquivos.

2764e99

Remove duplicidade de versão com e sem llama, assegura que sempre ser…

516c799

…á necessário um modelo llama.

Adiciona fixtures para testes.

e71b7e9

gitnnolabs requested review from Rossi-Luciano, pitangainnovare, robertatakenaka and samuelveigarangel November 30, 2025 14:17

gitnnolabs self-assigned this Nov 30, 2025

gitnnolabs added bug Something isn't working enhancement New feature or request labels Nov 30, 2025

robertatakenaka requested a review from Copilot December 8, 2025 14:37

Copilot started reviewing on behalf of robertatakenaka December 8, 2025 14:38 View session

Copilot AI reviewed Dec 8, 2025

View reviewed changes

robertatakenaka requested a review from Copilot March 24, 2026 12:22

Copilot started reviewing on behalf of robertatakenaka March 24, 2026 12:22 View session

Copilot AI reviewed Mar 24, 2026

View reviewed changes

robertatakenaka requested review from Copilot and eduranm March 24, 2026 12:44

Copilot started reviewing on behalf of robertatakenaka March 24, 2026 12:45 View session

Copilot AI reviewed Mar 24, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Correções de execução do markapi. #39

Correções de execução do markapi. #39
gitnnolabs wants to merge 6 commits intomarkupappfrom
fix-markupapp

gitnnolabs commented Nov 30, 2025

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Dec 8, 2025

Uh oh!

Copilot AI Dec 8, 2025

Uh oh!

Copilot AI Dec 8, 2025

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 24, 2026

Uh oh!

Copilot AI Mar 24, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 24, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

	LLAMA_ENABLED = env.bool("LLAMA_ENABLED", default=True)
	LLAMA_ENABLED = env.bool("LLAMA_ENABLED", default=False)

	estatus=1, # Estatus predeterminado
	status=1, # Estatus predeterminado

Conversation

gitnnolabs commented Nov 30, 2025

O que esse PR faz?

Onde a revisão poderia começar?

Como este poderia ser testado manualmente?

Algum cenário de contexto que queira dar?

Screenshots

Quais são tickets relevantes?

Referências

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Dec 8, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 8, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 8, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 24, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants