contrib: add per-chapter authors

.mailmap

@@ -0,0 +1 @@
+Biswaroop Bhattacharjee <[email protected]>

_config.yml

@@ -49,6 +49,7 @@ sphinx:
   extra_extensions:
   - sphinxemoji.sphinxemoji
   - sphinx_last_updated_by_git
+  - committers
   config:
     html_js_files:
     - [https://analytics.prem.ninja/js/script.js, {defer: defer, data-domain: book.premai.io}]

committers.py

@@ -0,0 +1,89 @@
+"""Print list of Git committer names.
+
+Example MyST usage (HTML only):
+
+    ```{committers} file_path.md
+    ```
+"""
+import json
+import os
+import re
+import subprocess
+from collections import Counter
+from datetime import timedelta
+from functools import cache
+from time import ctime, time
+
+from docutils import nodes
+from docutils.parsers.rst import Directive, directives
+from sphinx.application import Sphinx
+
+__version__ = '0.0.0'
+
+
+@cache
+def gh_user(email) -> str | None:
+    headers = [
+        '--header', 'Accept: application/vnd.github+json',
+        '--header', 'X-GitHub-Api-Version: 2022-11-28']
+    if (token := os.environ.get("GITHUB_TOKEN", os.environ.get("GH_TOKEN", ""))):
+        headers.extend(['--header', f'Authorization: Bearer {token}'])
+    for cmd in (
+        ['gh', 'api'] + headers + [f'search/users?q={email}+in:email'],
+        ['curl'] + headers + ['-fsSL', f'https://api.github.com/search/users?q={email}+in:email'],
+        ['wget'] + headers + ['-qO', '-', f'https://api.github.com/search/users?q={email}+in:email']
+    ):
+        try:
+            user_info = subprocess.check_output(cmd).decode('utf-8').strip()
+        except (subprocess.CalledProcessError, FileNotFoundError):
+            pass
+        else:
+            try:
+                return json.loads(user_info)['items'][0]['login']
+            except (KeyError, IndexError):
+                return
+
+
+class committers_node(nodes.General, nodes.Element):
+    pass
+
+
+def visit_committers_html(self, node):
+    self.body.append(self.starttag(node, 'div'))
+    self.body.append("Authors: ")
+    self.body.append(", ".join(f'<a href="{href}">{name}</a>' for name, href in node['authors']))
+
+
+def depart_committers_html(self, node):
+    self.body.append('</div>')
+
+
+class Committers(Directive):
+    has_content = True
+    required_arguments = 1
+    optional_arguments = 0
+    final_argument_whitespace = True
+    option_spec = {'class': directives.class_option, 'name': directives.unchanged}
+    _node = None
+
+    def run(self):
+        blame = subprocess.check_output([
+            'git', 'blame', '--line-porcelain', '-C2', '-M2', '--'] + self.arguments
+        ).decode('utf-8').strip()
+        authors = Counter(re.findall("^author (.*)\nauthor-mail <(.*)>", blame, flags=re.MULTILINE))
+        total_loc = authors.total()
+        auths = []
+        for (name, email), loc in authors.most_common():
+            if loc / total_loc < 0.05:
+                break
+            if (user := gh_user(email)):
+                auths.append((name, f"https://github.com/{user}"))
+            else:
+                auths.append((name, f"mailto:{email}"))
+        return [committers_node(authors=auths)]
+
+
+def setup(app: Sphinx):
+    app.add_node(committers_node, html=(visit_committers_html, depart_committers_html))
+    app.add_directive("committers", Committers)
+    return {'version': __version__, 'parallel_read_safe': True}

desktop-apps.md

@@ -10,3 +10,6 @@ See also:
 ## GPT4All UI
 
 {{ comments }}
+
+```{committers} desktop-apps.md
+```

eval-datasets.md

@@ -23,3 +23,6 @@ probably put short definition in [](glossary)?
 ## Reinforcement Learning for a Chat-based LLM
 
 {{ comments }}
+
+```{committers} eval-datasets.md
+```

fine-tuning.md

@@ -55,3 +55,6 @@ Falcon |
 ## Code Generators
 
 {{ comments }}
+
+```{committers} fine-tuning.md
+```

hardware.md

@@ -20,3 +20,6 @@ See also:
   + LLM-specific advice: https://gpus.llm-utils.org/cloud-gpu-guide/#which-gpu-cloud-should-i-use
 
 {{ comments }}
+
+```{committers} hardware.md
+```

licences.md

@@ -175,3 +175,6 @@ In the long term, we look forward to the outcomes of the US cases and EU proposa
 (licences-comments)=
 
 {{ comments }}
+
+```{committers} licences.md
+```

mlops-engines.md

@@ -27,3 +27,6 @@ See also https://onnxruntime.ai/docs/execution-providers
 ### Apache TVM
 
 {{ comments }}
+
+```{committers} mlops-engines.md
+```

model-formats.md

@@ -11,3 +11,6 @@ See also:
 ## TVM
 
 {{ comments }}
+
+```{committers} model-formats.md
+```

models.md

@@ -13,3 +13,6 @@ See also:
 - {cite}`golden-age-os-end`
 
 {{ comments }}
+
+```{committers} models.md
+```

sdk.md

@@ -9,3 +9,6 @@ See also: https://github.com/huggingface/transformers/blob/main/awesome-transfor
 ## Haystack
 
 {{ comments }}
+
+```{committers} sdk.md
+```

uncensored-models.md

@@ -5,3 +5,6 @@ FraudGPT, PoisonGPT
 "Cybercriminals train AI chatbots for phishing, malware attacks" (WormGPT: ChatGPT clone trained on malware-focused data, new: FraudGPT, coming soon: Bard-based version) https://www.bleepingcomputer.com/news/security/cybercriminals-train-ai-chatbots-for-phishing-malware-attacks
 
 {{ comments }}
+
+```{committers} uncensored-models.md
+```

vector-stores.md

@@ -11,3 +11,6 @@
 ## Chroma
 
 {{ comments }}
+
+```{committers} vector-stores.md
+```