[proofing] Add project stats

These statistics let us quantify the work required for a project, which
helps us with pricing projects for proofreaders.

Test plan: unit tests

ambuda/templates/macros/components.html

@@ -51,14 +51,15 @@
 
 {# A list of tabs. #}
 {% macro tabs(urls, active) %}
-<nav class="border-b mb-8">
-  <ul class="flex a-hover-underline">
+<nav class="md:border-b mb-8">
+  <ul class="border-l pl-4 md:border-0 md:p-0 md:flex a-hover-underline">
     {% for name, label, url in urls %}
     {% if name == active %}
-    <li class="p-4 font-semibold border-b-2 border-black">{{ label }}</li>
+    <li class="md:p-4 font-semibold underline 
+        md:no-underline md:border-b-2 md:border-black">{{ label }}</li>
     {% else %}
     <li class="font-semibold border-b-2 border-transparent text-slate-400">
-      <a class="block p-4" href="{{ url }}">{{ label }}</a>
+      <a class="block md:p-4" href="{{ url }}">{{ label }}</a>
     </li>
     {% endif %}
     {% endfor %}

ambuda/templates/macros/proofing.html

@@ -76,6 +76,7 @@ <h1 class="mt-2 font-bold">{{ _('Contribute') }}</h1>
   ("download", _("Download"),  url_for("proofing.project.download", slug=project.slug)),
 ] %}
 {% if is_mod %}{% set routes = routes + [
+  ("stats",    _("Stats"),     url_for("proofing.project.stats", slug=project.slug)),
   ("admin",    _("Admin"),     url_for("proofing.project.admin", slug=project.slug)),
 ] %}{% endif %}
 

ambuda/templates/proofing/projects/stats.html

@@ -0,0 +1,40 @@
+{% extends 'proofing/base-sidebar.html' %}
+{% import "macros/proofing.html" as m %}
+
+
+{% block title %}Stats: {{ project.title }} | Ambuda{% endblock %}
+
+
+{% block sidebar %}{{ m.main_nav('projects', current_user=current_user) }}{% endblock %}
+
+
+{% block content %}
+{{ m.project_header_nested('Stats', project) }}
+{{ m.project_tabs(project=project, active='stats', is_mod=current_user.is_moderator) }}
+
+<div class="prose">
+  <table>
+    <tr><th>&nbsp;</th><th>Total</th><th>Per page</th></tr>
+    <tr>
+        <th>Pages</th>
+        <td>{{ "{:,}".format(stats.num_pages) }}</td>
+        <td>&ndash;</td>
+    </tr>
+    <tr>
+        <th>Words</th>
+        <td>{{ "{:,}".format(stats.num_words) }}</td>
+        <td>{{ "{:,}".format(stats.num_words // stats.num_pages) }}</td>
+    </tr>
+    <tr>
+        <th>Roman characters</th>
+        <td>{{ "{:,}".format(stats.num_roman_characters) }}</td>
+        <td>{{ "{:,}".format(stats.num_roman_characters // stats.num_pages) }}</td></tr>
+    </tr>
+    <tr>
+        <th>Aksharas</th>
+        <td>{{ "{:,}".format(stats.num_aksharas) }}</td>
+        <td>{{ "{:,}".format(stats.num_aksharas // stats.num_pages) }}</td>
+    </tr>
+  </table>
+</div>
+{% endblock %}

ambuda/views/proofing/project.py

@@ -37,6 +37,7 @@
 from ambuda.utils import project_utils, proofing_utils
 from ambuda.utils.revisions import add_revision
 from ambuda.views.proofing.decorators import moderator_required, p2_required
+from ambuda.views.proofing.stats import calculate_stats
 
 bp = Blueprint("project", __name__)
 LOG = logging.getLogger(__name__)
@@ -275,6 +276,24 @@ def download_as_xml(slug):
     return response
 
 
+@bp.route("/<slug>/stats")
+@moderator_required
+def stats(slug):
+    """Show basic statistics about this project.
+
+    Currently, these stats don't show any sensitive information. But since that
+    might change in the future, limit this page to moderators only.
+    """
+    project_ = q.project(slug)
+    if project_ is None:
+        abort(404)
+
+    stats_ = calculate_stats(project_)
+    return render_template(
+        "proofing/projects/stats.html", project=project_, stats=stats_
+    )
+
+
 @bp.route("/<slug>/search")
 @login_required
 def search(slug):

ambuda/views/proofing/stats.py

@@ -0,0 +1,81 @@
+"""Utilities for calculating project statistics.
+
+These stats are useful for estimating the size of a project, which helps us
+wtih cost estimation for our proofers.
+"""
+
+import re
+from collections.abc import Iterable
+from dataclasses import dataclass
+
+from indic_transliteration import detect, sanscript
+
+from ambuda.database import Project
+
+#: Matches whitespace spans.
+RE_SPACE = re.compile(r"\s+", re.MULTILINE)
+#: Matches Sanskrit vowels in SLP1.
+RE_VOWEL = re.compile(r"[aAiIuUfFxXeEoO]")
+
+#: Scripts that don't use aksharas
+#: (copied from `indic_transliteration`)
+ROMAN_SCHEMES = {
+    "hk",
+    "iast",
+    "itrans",
+    "kolkata_v2",
+    "slp1",
+    "velthuis",
+}
+
+
+@dataclass
+class Stats:
+    """Statistics for some project."""
+
+    #: The number of pages.
+    num_pages: int
+    #: The number of words.
+    #: Here, a "word" is a continuous span of characters with no whitespace.
+    num_words: int
+    #: The number of Roman characters.
+    num_roman_characters: int
+    #: The number of aksharas (syllables) in Devanagari, Kannada, or some other
+    #: Brahmic script.
+    num_aksharas: int
+
+
+def _calculate_stats_for_strings(strings: Iterable[str]) -> Stats:
+    num_pages = 0
+    num_words = 0
+    num_roman_characters = 0
+    num_aksharas = 0
+    for page_text in strings:
+        num_pages += 1
+        # N words will have n-1 spaces, so add 1 to get a better word count.
+        spaces = RE_SPACE.findall(page_text)
+        num_words += 1 + len(spaces)
+
+        encoding = detect.detect(page_text)
+        if encoding in ROMAN_SCHEMES:
+            num_space_chars = sum(len(x) for x in spaces)
+            num_roman_characters += len(page_text) - num_space_chars
+        else:
+            slp1_text = sanscript.transliterate(page_text, encoding, "slp1")
+            num_aksharas += len(RE_VOWEL.findall(slp1_text))
+
+    return Stats(
+        num_pages=num_pages,
+        num_words=num_words,
+        num_roman_characters=num_roman_characters,
+        num_aksharas=num_aksharas,
+    )
+
+
+def _iter_page_strings(project: Project) -> Iterable[str]:
+    for page in project.pages:
+        yield page.revisions[-1].content if page.revisions else ""
+
+
+def calculate_stats(project: Project) -> Stats:
+    return _calculate_stats_for_strings(_iter_page_strings(project))

test/ambuda/views/proofing/test_project.py

@@ -98,6 +98,15 @@ def test_download_as_xml__bad_project(client):
     assert resp.status_code == 404
 
 
+def test_stats(moderator_client, rama_client):
+    resp = moderator_client.get("/proofing/test-project/stats")
+    assert resp.status_code == 200
+    assert "Roman characters" in resp.text
+
+    resp = rama_client.get("/proofing/test-project/stats")
+    assert resp.status_code == 302
+
+
 def test_search(rama_client):
     resp = rama_client.get("/proofing/test-project/search")
     assert "Search:" in resp.text

test/ambuda/views/proofing/test_stats.py

@@ -0,0 +1,16 @@
+from ambuda.views.proofing.stats import _calculate_stats_for_strings
+
+
+def test_calculate_stats():
+    strings = [
+        "Test",
+        "अहम्",
+        "astyuttarasyāṃ diśi devatātmā",
+        "foo bar baz",
+    ]
+    stats = _calculate_stats_for_strings(strings)
+    assert stats.num_pages == 4
+    assert stats.num_words == 8
+    assert stats.num_aksharas == 2
+    # 4 + (29 - 2) + (11 - 2) = 40
+    assert stats.num_roman_characters == 40