photo-stack-finder 0.1.7__py3-none-any.whl → 0.1.8__py3-none-any.whl

utils/template_parsing.py

@@ -0,0 +1,71 @@
+"""Template parsing utilities for filename pattern extraction.
+
+This module provides utilities for extracting template patterns from filenames
+by replacing digit sequences with placeholders. This enables grouping files
+that follow the same naming pattern (e.g., IMG_0001.jpg, IMG_0002.jpg).
+
+The template extraction happens early in the pipeline (during SHA binning) and
+is cached on PhotoFile objects for use in later stages.
+"""
+
+from __future__ import annotations
+
+import re
+
+# Type alias for digit sequence indices extracted from filenames
+# Example: "IMG_1234_5678.jpg" → ("1234", "5678")
+INDEX_T = tuple[str, ...]
+
+
+def extract_template(text: str) -> tuple[str, INDEX_T]:
+    """Extract template and digit groups from text.
+
+    Replaces all digit sequences with placeholders {P0}, {P1}, etc.
+    This enables grouping files with similar naming patterns.
+
+    The function scans the input text for consecutive digit sequences and
+    replaces each with a placeholder of the form {P0}, {P1}, etc. The
+    original digit sequences are returned as a tuple for later reconstruction.
+
+    Args:
+        text: Text to parse (typically filename stem without extension)
+
+    Returns:
+        Tuple of (template_string, digit_sequences) where:
+        - template_string: Text with digit sequences replaced by {P0}, {P1}, etc.
+        - digit_sequences: Tuple of original digit strings in order of appearance
+
+    Examples:
+        >>> template, digits = extract_template("IMG_1234")
+        >>> print(template)
+        IMG_{P0}
+        >>> print(digits)
+        ('1234',)
+
+        >>> template, digits = extract_template("2023_05_IMG_1234_5678")
+        >>> print(template)
+        {P0}_{P1}_IMG_{P2}_{P3}
+        >>> print(digits)
+        ('2023', '05', '1234', '5678')
+
+        >>> template, digits = extract_template("no_digits_here")
+        >>> print(template)
+        no_digits_here
+        >>> print(digits)
+        ()
+
+        >>> template, digits = extract_template("12345")
+        >>> print(template)
+        {P0}
+        >>> print(digits)
+        ('12345',)
+    """
+    digits: list[str] = []
+
+    def repl(m: re.Match[str]) -> str:
+        """Replace matched digit sequence with placeholder {PN} and capture digits."""
+        digits.append(m.group(0))
+        return f"{{P{len(digits) - 1}}}"
+
+    template: str = re.sub(r"\d+", repl, text)
+    return template, tuple(digits)

photo_stack_finder-0.1.7.dist-info/RECORD

@@ -1,74 +0,0 @@
-orchestrator/__init__.py,sha256=DtoeLEhNbKfBp-JuFx-4DUFrLFrO8HUV0lSov1sYfxE,1315
-orchestrator/app.py,sha256=ZqlJ_ZE0gC2P7Nd3mGNu9jMwENQOboNoRV5ZRfCM0BM,34302
-orchestrator/build_pipeline.py,sha256=MTkrG5nV_a3xe8P4doKWRN2-_4ew5HzsFJHfKuChytk,4385
-orchestrator/orchestrator_runner.py,sha256=dV_5cuQxZvXiSkj7s0Y_ScD5u2af0UusvuksBXjmOsI,20062
-orchestrator/pipeline_builder.py,sha256=R-Y92GjDsBYz5wX5jYgzeKUmXEWjSJIRJ5mj6oDHQ6U,4675
-orchestrator/pipeline_orchestrator.py,sha256=vvqI85TZNHSaxeCDeAn3oSZl8BObORp9ofMhg9ZoTAo,25932
-orchestrator/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-orchestrator/review_persistence.py,sha256=MGRTEmEVfVQNbo_VycXW98a92OeexyS-TmGN8vBiTVA,5913
-orchestrator/static/favicon.svg,sha256=TVNER-nLkPBBb9lZRaruak9jvyXyyfIKgyD2koQcyqk,622
-orchestrator/static/orchestrator.css,sha256=PKeb4qYWpXukbUfZuVeNqQnQc1OlqdwPUGIkEqq77Jo,11590
-orchestrator/static/orchestrator.html,sha256=6OmxtX3Un-1Ru6sWtL_I49Q4gplwuwG07hXYrTMvCDc,13510
-orchestrator/static/orchestrator.js,sha256=Hz5DPEwl0FbCTMhLDOwFHvxAWCN4MGcb1FVzsUyaPlY,32313
-orchestrator/static/review_common.js,sha256=t8-AI4ohQbtqSC1iWvR5x0FNlwIxl8gVO2qZsI-VtnI,5875
-orchestrator/static/review_identical.html,sha256=Pq8TYbDSBbajC3p_XvlIOC1lbv0eJN_TcymGUw1lDE8,31487
-orchestrator/static/review_sequences.html,sha256=oEWRtcMfEvu09F3r-hw8rRfECjcE48BHtht2CdGlw3E,54681
-overlap_metrics/__init__.py,sha256=uUZORsoNBeLrl9w7eNtsaz-mEl_dX8oB1Ivp4hI1l6E,11185
-overlap_metrics/config.py,sha256=kfCIVJpE4TLrrjeLh4Dk4BIYw8zVqYzCyCvhXMMYN40,3967
-overlap_metrics/core.py,sha256=a3s4_NR8nvVp3HVmXhCppG5tIZsTMciplp9UYR08YnM,9656
-overlap_metrics/estimators.py,sha256=qVHGpKi7HaLXVI5J0dJjl5Ml0dejP8Vfv048M9PUJFg,11033
-overlap_metrics/metrics.py,sha256=eo9DtQRG2QTkKXYeXaXybJMyysHteQwzVrNoIPpbcA8,10180
-overlap_metrics/registry.py,sha256=bDgakoh_sxDOv2wMh_-iFgIKCe7td-kOGHwo5hnQMGE,3651
-overlap_metrics/utils.py,sha256=0WQxaRfjtxmEOjnJHXsq3PkWhxICkwoY-jG9sNqbBLY,4625
-photo_compare/__init__.py,sha256=g8e1wex31_Ku5Lry75xOlFSzRLPqCSPBPxgiLwsxer0,4877
-photo_compare/base.py,sha256=oxBR6VEwOXjw8_Jd6PgQtiv4qHpxU4zMRDl4QMOnnv4,9641
-photo_compare/config.py,sha256=qrhJvmx9Io9gxlZWB7GHynBxTGBep8Y9SoZXM2sIviY,7781
-photo_compare/distance.py,sha256=fcsitsn9MyA2o4UhMz5PguIGyJAbeh3wXFgKdvQp2d4,581
-photo_compare/feature_methods.py,sha256=zwKkPP1q8jw4iVUtTTPn3CnJOZXiCjMQL0lL3rxjNC8,7077
-photo_compare/file_hash.py,sha256=dr42d3XLMJ6rjCXSwWXLknpZ__PLiOC-Bx3t7tPugeM,1006
-photo_compare/hash_methods.py,sha256=yJGj8dqJ-ugguptKIarM9ot9dzbiHKkErwg6BnQArbc,3783
-photo_compare/histogram_methods.py,sha256=QVKCWJ2rLp52mMVJphbEJ1RCN935hPQSWXaRwnn7t9Q,5136
-photo_compare/pixel_methods.py,sha256=9uOl2qRiPrNgNNcP0B1a21EGHmSA9SUXguegmg0bVME,2421
-photo_compare/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-photo_compare/structural_methods.py,sha256=GGigVlxLgApHg2Zj4x9kZf-9IBymQMkkRR_yIeG0qHc,4359
-photo_compare/types.py,sha256=oru4zQdA43FhCtgV8BSiJrsS7gIuVtRoaDxh_BLaOtQ,999
-photo_stack_finder-0.1.7.dist-info/licenses/LICENSE,sha256=bmBV3PYoSFRs600gmAehqmuaIZIVUl8kGQoZWhapVnM,1468
-scripts/orchestrate.py,sha256=DYcBTxaJLzMGE4KAwsyimcXlw4nPgN3BTQNdnFeKxSs,9215
-utils/__init__.py,sha256=nCUhcOaBdOmc57VwLwRNrzzQqxPAjNI6H01-kKvoGuE,5141
-utils/base_pipeline_stage.py,sha256=PbYBJ25Q8gPHvLwbtaJXkVyiLT99QP9l4b_W3CyBlnc,6540
-utils/base_ports.py,sha256=WbR38BlcAkbQaOtX06yDd2pDOx09pB0Mh4OkWvcxfEY,5416
-utils/benchmark_utils.py,sha256=aBt_LyrCXgO3Z98bX9DXmpJW-HwhSEs43MZyoHxrL5M,30849
-utils/channel.py,sha256=cSmP2y5_vS0YNHPm1CFOTACGeeTxwfbv0u5y_AfhIrw,2660
-utils/comparison_gates.py,sha256=FMUCQq4tHEx5WGpTgdq5bHYatI-VjIEcDoFA-DU2n5s,23468
-utils/compute_benchmarks.py,sha256=gnqMdgecP23AmlK7mdGrdHLd1S7Ks3zm6-UvuKUkgSs,14223
-utils/compute_identical.py,sha256=3a6GiwInaEglduy_dn-8zMhlePf7oJ9VeIvQ5Bw5BqY,7111
-utils/compute_indices.py,sha256=1iYTgl9uTVY8k2xaSXSKgR1voDK88LRCAqdCPlMYKX4,9539
-utils/compute_perceptual_hash.py,sha256=DP_19r0sBDHcsDIsPoGGFtalbSmyjCJpbti7c6i7DhI,5042
-utils/compute_perceptual_match.py,sha256=1XBmBt0H8IeeITDUjjOgpm-Y3Eu7AwZEi4hecl9qIzI,10498
-utils/compute_sha_bins.py,sha256=yjPoYPfl-BV9Wo8HM8pCZe_Abxw1sztuWoKM2uuUClA,5423
-utils/compute_template_similarity.py,sha256=n_5ogh-0icwT01YCqgX4wXEzp8HTgBYCthY8ceR_GX0,21448
-utils/compute_templates.py,sha256=yVYfmaJlfOgeDJ2tdJ5iLd-n4SnPbjepl7eUBP3N1Js,5079
-utils/compute_versions.py,sha256=5ihwycoNo4uf8YsVmfEvmWJN1hOxCA-1HXDFS7pQjP0,20819
-utils/config.py,sha256=TRPFe_4-20hXkNS3lnGHTVv7vuyp3fxmoKqg12jsCVQ,15485
-utils/data_io.py,sha256=S1lJxGLa-uxLYNrIwf3Md9_672jA18U98l24qOYkIQM,2495
-utils/graph_context.py,sha256=ubosGjtE0XrADw5x9CYxJ6cebynqhGQ7dcetZy8ZXpY,1602
-utils/logger.py,sha256=AaidxSlr4S1wr2J5lXo3hV4fEAht2skzft5yFVJdzuY,1873
-utils/models.py,sha256=oaWVyLRMSM7hKMjkaGYXaFWC0uAqqnUHQBNp6Gj0JO4,15300
-utils/photo_file.py,sha256=uxX37yiuzqUQ32jIuaqE6LyaRv-WqWiRyrzmeL3nXGI,21411
-utils/pipeline_graph.py,sha256=-w3N43_GSQUb21lYVDSxmi-y6bbLF98m5hdduka1eQk,13596
-utils/pipeline_stage.py,sha256=iI7pluzUxgjqGW7J822sR8J77iYdOYWWqQODQWXXUBw,17386
-utils/plot_helpers.py,sha256=H3Ll4IfocUITLrDW8P_qKuhUrbJN4213ZpBQA41OJwQ,4574
-utils/ports.py,sha256=45yTUzoghSFb2uHq05lAmlLkOIM-TBt3e4boBjpJaOA,4454
-utils/progress.py,sha256=lH6aorgRxz-cYmkrIS3L8OKxH_Na9Nq59alrXKGphFs,15351
-utils/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-utils/report_builder.py,sha256=eYxrYym6uZj719dKMkGm-eDrUNg8zEosn20hhLsSI6I,4079
-utils/review_types.py,sha256=UDzyPKrm-6ICMKo51YbQUJWolFK01mBEbNS0ziocYi8,1768
-utils/review_utils.py,sha256=CA4vSVrHTKB8-v7Xj5OJk4u1bcb3qCnIVspZVkWo4IA,8999
-utils/sequence.py,sha256=WTM8L8if4KD2yqXhQFnJuNXeRFpF39vAiG1fU2IaBTQ,34879
-utils/sequence_clustering.py,sha256=QyqYXHvnzuoknjpyImOFB9D4qEk7WDekP-C4kjXaPqw,5754
-utils/template.py,sha256=pRJqlf4QyzdHC0ea3L0o_sAVq_awtP34nYBvYhL2Vb8,1761
-photo_stack_finder-0.1.7.dist-info/METADATA,sha256=sPXgLHR4hvS7hy-urNKNikPRo-FuQmY3lJbrgAGiz2U,17117
-photo_stack_finder-0.1.7.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-photo_stack_finder-0.1.7.dist-info/entry_points.txt,sha256=ZHWJIeylUGZYa_8Rjt3QxbWiCCVZZx3FQShTNg91zNQ,64
-photo_stack_finder-0.1.7.dist-info/top_level.txt,sha256=VmNqTeDiNgCj3LZJICyRF2Uc4buZQ7LxgSob-BoImk0,57
-photo_stack_finder-0.1.7.dist-info/RECORD,,