phykit 2.1.4__tar.gz → 2.1.8__tar.gz

{phykit-2.1.4 → phykit-2.1.8}/PKG-INFO

@@ -1,18 +1,17 @@
 Metadata-Version: 2.1
 Name: phykit
-Version: 2.1.4
+Version: 2.1.8
 Home-page: https://github.com/jlsteenwyk/phykit
 Author: Jacob L. Steenwyk
 Author-email: jlsteenwyk@gmail.com
 Classifier: Operating System :: OS Independent
 Classifier: Intended Audience :: Science/Research
 Classifier: Programming Language :: Python
-Classifier: Programming Language :: Python :: 3.9
-Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Scientific/Engineering
+Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 License-File: LICENSE.md
 
@@ -34,8 +33,8 @@ License-File: LICENSE.md
         <a href="https://codecov.io/gh/jlsteenwyk/phykit" alt="Coverage">
           <img src="https://codecov.io/gh/jlsteenwyk/phykit/branch/master/graph/badge.svg?token=0J49I6441V">
         </a>
-        <a href="https://github.com/jlsteenwyk/phykit/graphs/contributors" alt="Contributors">
-            <img src="https://img.shields.io/github/contributors/jlsteenwyk/phykit">
+        <a href="https://github.com/JLSteenwyk/PhyKIT/graphs/contributors" alt="Contributors">
+            <img src="https://img.shields.io/github/contributors/JLSteenwyk/PhyKIT">
         </a>
         <a href="https://bsky.app/profile/jlsteenwyk.bsky.social" target="_blank" rel="noopener noreferrer">
           <img src="https://img.shields.io/badge/Bluesky-0285FF?logo=bluesky&logoColor=fff">

{phykit-2.1.4 → phykit-2.1.8}/README.md

@@ -16,8 +16,8 @@
         <a href="https://codecov.io/gh/jlsteenwyk/phykit" alt="Coverage">
           <img src="https://codecov.io/gh/jlsteenwyk/phykit/branch/master/graph/badge.svg?token=0J49I6441V">
         </a>
-        <a href="https://github.com/jlsteenwyk/phykit/graphs/contributors" alt="Contributors">
-            <img src="https://img.shields.io/github/contributors/jlsteenwyk/phykit">
+        <a href="https://github.com/JLSteenwyk/PhyKIT/graphs/contributors" alt="Contributors">
+            <img src="https://img.shields.io/github/contributors/JLSteenwyk/PhyKIT">
         </a>
         <a href="https://bsky.app/profile/jlsteenwyk.bsky.social" target="_blank" rel="noopener noreferrer">
           <img src="https://img.shields.io/badge/Bluesky-0285FF?logo=bluesky&logoColor=fff">

{phykit-2.1.4 → phykit-2.1.8}/phykit/__main__.py

@@ -1,5 +1,4 @@
 """phykit.__main__: executed when phykit is called as script"""
-import sys
 
 from .phykit import Phykit
 

phykit-2.1.8/phykit/cli_registry.py

@@ -0,0 +1,105 @@
+"""Declarative CLI alias registry."""
+
+from typing import Dict
+
+
+# Map user-facing aliases (and a few long-form synonyms) to Phykit handler names.
+ALIAS_TO_HANDLER: Dict[str, str] = {
+    # version
+    "version": "version",
+    "v": "version",
+    # Alignment aliases
+    "aln_len": "alignment_length",
+    "al": "alignment_length",
+    "aln_len_no_gaps": "alignment_length_no_gaps",
+    "alng": "alignment_length_no_gaps",
+    "aln_entropy": "alignment_entropy",
+    "entropy": "alignment_entropy",
+    "aln_recoding": "alignment_recoding",
+    "recode": "alignment_recoding",
+    "outlier_taxa": "alignment_outlier_taxa",
+    "aot": "alignment_outlier_taxa",
+    "cs": "column_score",
+    "comp_bias_per_site": "compositional_bias_per_site",
+    "cbps": "compositional_bias_per_site",
+    "evo_rate_per_site": "evolutionary_rate_per_site",
+    "erps": "evolutionary_rate_per_site",
+    "get_entry": "faidx",
+    "ge": "faidx",
+    "gc": "gc_content",
+    "mask_aln": "mask_alignment",
+    "mask": "mask_alignment",
+    "plot_qc": "plot_alignment_qc",
+    "paqc": "plot_alignment_qc",
+    "occupancy_taxon": "occupancy_per_taxon",
+    "occ_tax": "occupancy_per_taxon",
+    "pairwise_id": "pairwise_identity",
+    "pi": "pairwise_identity",
+    "comp_taxon": "composition_per_taxon",
+    "comp_tax": "composition_per_taxon",
+    "pis": "parsimony_informative_sites",
+    "rel_comp_var": "rcv",
+    "relative_composition_variability": "rcv",
+    "relative_composition_variability_taxon": "rcvt",
+    "rel_comp_var_taxon": "rcvt",
+    "rename_fasta": "rename_fasta_entries",
+    "sum_of_pairs_score": "sum_of_pairs_score",
+    "sops": "sum_of_pairs_score",
+    "sop": "sum_of_pairs_score",
+    "vs": "variable_sites",
+    # Tree aliases
+    "bss": "bipartition_support_stats",
+    "blm": "branch_length_multiplier",
+    "collapse": "collapse_branches",
+    "cb": "collapse_branches",
+    "cover": "covarying_evolutionary_rates",
+    "consensus": "consensus_tree",
+    "ctree": "consensus_tree",
+    "degree_of_violation_of_a_molecular_clock": "dvmc",
+    "evo_rate": "evolutionary_rate",
+    "clan_check": "hidden_paralogy_check",
+    "ibs": "internal_branch_stats",
+    "il": "internode_labeler",
+    "lca_subtree": "last_common_ancestor_subtree",
+    "long_branch_score": "lb_score",
+    "lbs": "lb_score",
+    "is_monophyletic": "monophyly_check",
+    "nni": "nearest_neighbor_interchange",
+    "pd": "patristic_distances",
+    "phylo_signal": "phylogenetic_signal",
+    "ps": "phylogenetic_signal",
+    "polyt_test": "polytomy_test",
+    "ptt": "polytomy_test",
+    "polyt": "polytomy_test",
+    "print": "print_tree",
+    "pt": "print_tree",
+    "prune": "prune_tree",
+    "rename_tree": "rename_tree_tips",
+    "rename_tips": "rename_tree_tips",
+    "robinson_foulds_distance": "rf_distance",
+    "rf_dist": "rf_distance",
+    "rf": "rf_distance",
+    "root": "root_tree",
+    "rt": "root_tree",
+    "spurious_seq": "spurious_sequence",
+    "ss": "spurious_sequence",
+    "tbs": "terminal_branch_stats",
+    "labels": "tip_labels",
+    "tree_labels": "tip_labels",
+    "tl": "tip_labels",
+    "t2t_dist": "tip_to_tip_distance",
+    "t2t": "tip_to_tip_distance",
+    "t2t_node_dist": "tip_to_tip_node_distance",
+    "t2t_nd": "tip_to_tip_node_distance",
+    "tree_len": "total_tree_length",
+    "tness": "treeness",
+    # Alignment- and tree-based aliases
+    "sat": "saturation",
+    "toverr": "treeness_over_rcv",
+    "tor": "treeness_over_rcv",
+    # Helper aliases
+    "create_concat": "create_concatenation_matrix",
+    "cc": "create_concatenation_matrix",
+    "pal2nal": "thread_dna",
+    "p2n": "thread_dna",
+}

phykit-2.1.8/phykit/errors.py

@@ -0,0 +1,12 @@
+"""Shared PhyKIT exception types."""
+
+from typing import Iterable, List
+
+
+class PhykitUserError(SystemExit):
+    """User-facing error that should be rendered by the CLI with a specific exit code."""
+
+    def __init__(self, messages: Iterable[str], code: int = 2):
+        self.messages: List[str] = list(messages)
+        super().__init__(code)
+

{phykit-2.1.4 → phykit-2.1.8}/phykit/helpers/caching.py

@@ -57,7 +57,7 @@ class ResultCache:
             try:
                 with open(cache_file, 'rb') as f:
                     return pickle.load(f)
-            except:
+            except Exception:
                 # Cache corrupted, remove it
                 os.remove(cache_file)
 
@@ -70,7 +70,7 @@ class ResultCache:
         try:
             with open(cache_file, 'wb') as f:
                 pickle.dump(value, f)
-        except:
+        except Exception:
             # Caching failed, continue without caching
             pass
 
@@ -162,6 +162,8 @@ class AlignmentCache:
         Cache alignment column.
         """
         self._column_cache[f"{alignment_hash}_{column_idx}"] = column
+        # Prevent stale values from the lru_cache layer.
+        self.get_column.cache_clear()
 
     @lru_cache(maxsize=128)
     def get_stats(self, alignment_hash: str, stat_type: str) -> Any:
@@ -175,6 +177,8 @@ class AlignmentCache:
         Cache alignment statistics.
         """
         self._stats_cache[f"{alignment_hash}_{stat_type}"] = stats
+        # Prevent stale values from the lru_cache layer.
+        self.get_stats.cache_clear()
 
     def clear(self) -> None:
         """
@@ -198,4 +202,4 @@ def get_result_cache() -> ResultCache:
 
 def get_alignment_cache() -> AlignmentCache:
     """Get global alignment cache instance."""
-    return _alignment_cache
+    return _alignment_cache

{phykit-2.1.4 → phykit-2.1.8}/phykit/helpers/files.py

@@ -1,5 +1,4 @@
 from enum import Enum
-import sys
 from typing import Tuple, Optional
 from functools import lru_cache
 import hashlib
@@ -7,6 +6,7 @@ import os
 
 from Bio import AlignIO
 from Bio.Align import MultipleSeqAlignment
+from ..errors import PhykitUserError
 
 
 class FileFormat(Enum):
@@ -30,7 +30,7 @@ def _get_file_hash(file_path: str) -> str:
 
 def _detect_format_by_content(file_path: str) -> Optional[str]:
     """Attempt to detect file format by examining file content."""
-    with open(file_path, 'r') as f:
+    with open(file_path) as f:
         first_line = f.readline().strip()
 
         # Quick format detection based on first line
@@ -60,9 +60,13 @@ def get_alignment_and_format(
 ) -> Tuple[MultipleSeqAlignment, str, bool]:
     # Check if file exists first
     if not os.path.exists(alignment_file_path):
-        print(f"{alignment_file_path} corresponds to no such file.")
-        print("Please check file name and pathing")
-        sys.exit(2)
+        raise PhykitUserError(
+            [
+                f"{alignment_file_path} corresponds to no such file.",
+                "Please check file name and pathing",
+            ],
+            code=2,
+        )
 
     # Try to detect format by content first
     detected_format = _detect_format_by_content(alignment_file_path)
@@ -95,9 +99,13 @@ def get_alignment_and_format(
             continue
 
     # If we get here, no format worked
-    print(f"Could not determine format for {alignment_file_path}")
-    print("Please ensure the file is in a supported format")
-    sys.exit(2)
+    raise PhykitUserError(
+        [
+            f"Could not determine format for {alignment_file_path}",
+            "Please ensure the file is in a supported format",
+        ],
+        code=2,
+    )
 
 
 def is_protein_alignment(alignment: MultipleSeqAlignment) -> bool:
@@ -120,6 +128,10 @@ def read_single_column_file_to_list(single_col_file_path: str) -> list:
         with open(single_col_file_path) as f:
             return [line.rstrip("\n").strip() for line in f]
     except FileNotFoundError:
-        print(f"{single_col_file_path} corresponds to no such file or directory.")
-        print("Please check file name and pathing")
-        sys.exit(2)
+        raise PhykitUserError(
+            [
+                f"{single_col_file_path} corresponds to no such file or directory.",
+                "Please check file name and pathing",
+            ],
+            code=2,
+        )

phykit-2.1.8/phykit/helpers/json_output.py

@@ -0,0 +1,26 @@
+import json
+
+import numpy as np
+
+
+def to_builtin_json_types(value):
+    if isinstance(value, dict):
+        return {key: to_builtin_json_types(sub_value) for key, sub_value in value.items()}
+    if isinstance(value, list):
+        return [to_builtin_json_types(sub_value) for sub_value in value]
+    if isinstance(value, tuple):
+        return [to_builtin_json_types(sub_value) for sub_value in value]
+    if isinstance(value, np.integer):
+        return int(value)
+    if isinstance(value, np.floating):
+        return float(value)
+    if isinstance(value, np.ndarray):
+        return [to_builtin_json_types(sub_value) for sub_value in value.tolist()]
+    return value
+
+
+def print_json(payload, sort_keys=True):
+    try:
+        print(json.dumps(to_builtin_json_types(payload), sort_keys=sort_keys))
+    except BrokenPipeError:
+        pass

{phykit-2.1.4 → phykit-2.1.8}/phykit/helpers/parallel.py

@@ -3,7 +3,6 @@ Parallel processing utilities for batch operations
 """
 
 import multiprocessing as mp
-from functools import partial
 from typing import List, Any, Callable, Optional, Tuple
 import numpy as np
 from concurrent.futures import ProcessPoolExecutor, ThreadPoolExecutor, as_completed
@@ -302,4 +301,4 @@ class NumpyParallel:
             if symmetric:
                 result_matrix[j, i] = value
 
-        return result_matrix
+        return result_matrix

{phykit-2.1.4 → phykit-2.1.8}/phykit/helpers/streaming.py

@@ -2,7 +2,7 @@
 Streaming utilities for memory-efficient processing of large files
 """
 
-from typing import Iterator, Tuple, Optional
+from typing import Iterator, Optional
 import mmap
 import os
 from Bio import SeqIO
@@ -31,9 +31,8 @@ class StreamingFastaReader:
         """
         Stream sequences one at a time.
         """
-        with open(self.file_path, 'r') as handle:
-            for record in SeqIO.parse(handle, "fasta"):
-                yield record
+        with open(self.file_path) as handle:
+            yield from SeqIO.parse(handle, "fasta")
 
     def stream_chunks(self) -> Iterator[list]:
         """
@@ -149,4 +148,4 @@ class MemoryEfficientAlignmentProcessor:
             batch_result = processing_func(batch)
             results.append(batch_result)
 
-        return results
+        return results