logdetective 1.4.0__py3-none-any.whl → 1.6.0__py3-none-any.whl

logdetective/constants.py

@@ -26,17 +26,6 @@ Analysis:
 
 """
 
-SUMMARIZATION_PROMPT_TEMPLATE = """
-Does following log contain error or issue?
-
-Log:
-
-{}
-
-Answer:
-
-"""
-
 SNIPPET_PROMPT_TEMPLATE = """
 Analyse following RPM build log snippet. Describe contents accurately, without speculation or suggestions for resolution.
 

logdetective/extractors.py

@@ -4,79 +4,23 @@ from typing import Tuple
 
 import drain3
 from drain3.template_miner_config import TemplateMinerConfig
-from llama_cpp import Llama, LlamaGrammar
 
-from logdetective.constants import SUMMARIZATION_PROMPT_TEMPLATE
-from logdetective.utils import get_chunks
+from logdetective.utils import get_chunks, filter_snippet_patterns
+from logdetective.models import SkipSnippets
 
 LOG = logging.getLogger("logdetective")
 
 
-class LLMExtractor:
-    """
-    A class that extracts relevant information from logs using a language model.
-    """
+class DrainExtractor:
+    """A class that extracts information from logs using a template miner algorithm."""
 
     def __init__(
         self,
-        model: Llama,
-        n_lines: int = 2,
-        prompt: str = SUMMARIZATION_PROMPT_TEMPLATE,
+        verbose: bool = False,
+        context: bool = False,
+        max_clusters=8,
+        skip_snippets: SkipSnippets = SkipSnippets({}),
     ):
-        self.model = model
-        self.n_lines = n_lines
-        self.grammar = LlamaGrammar.from_string(
-            'root ::= ("Yes" | "No")', verbose=False
-        )
-        self.prompt = prompt
-
-    def __call__(
-        self, log: str, n_lines: int = 2, neighbors: bool = False
-    ) -> list[str]:
-        chunks = self.rate_chunks(log)
-        out = self.create_extract(chunks, neighbors)
-        return out
-
-    def rate_chunks(self, log: str) -> list[tuple]:
-        """Scan log by the model and store results.
-
-        :param log: log file content
-        """
-        results = []
-        log_lines = log.split("\n")
-
-        for i in range(0, len(log_lines), self.n_lines):
-            block = "\n".join(log_lines[i: i + self.n_lines])
-            prompt = self.prompt.format(log)
-            out = self.model(prompt, max_tokens=7, grammar=self.grammar)
-            out = f"{out['choices'][0]['text']}\n"
-            results.append((block, out))
-
-        return results
-
-    def create_extract(self, chunks: list[tuple], neighbors: bool = False) -> list[str]:
-        """Extract interesting chunks from the model processing."""
-        interesting = []
-        summary = []
-        # pylint: disable=consider-using-enumerate
-        for i in range(len(chunks)):
-            if chunks[i][1].startswith("Yes"):
-                interesting.append(i)
-                if neighbors:
-                    interesting.extend([max(i - 1, 0), min(i + 1, len(chunks) - 1)])
-
-        interesting = set(interesting)
-
-        for i in interesting:
-            summary.append(chunks[i][0])
-
-        return summary
-
-
-class DrainExtractor:
-    """A class that extracts information from logs using a template miner algorithm."""
-
-    def __init__(self, verbose: bool = False, context: bool = False, max_clusters=8):
         config = TemplateMinerConfig()
         config.load(f"{os.path.dirname(__file__)}/drain3.ini")
         config.profiling_enabled = verbose
@@ -84,22 +28,28 @@ class DrainExtractor:
         self.miner = drain3.TemplateMiner(config=config)
         self.verbose = verbose
         self.context = context
+        self.skip_snippets = skip_snippets
 
     def __call__(self, log: str) -> list[Tuple[int, str]]:
         out = []
+        # Create chunks
+        chunks = list(get_chunks(log))
+        # Keep only chunks that don't match any of the excluded patterns
+        chunks = [
+            (_, chunk)
+            for _, chunk in chunks
+            if not filter_snippet_patterns(chunk, self.skip_snippets)
+        ]
         # First pass create clusters
-        for _, chunk in get_chunks(log):
+        for _, chunk in chunks:
             processed_chunk = self.miner.add_log_message(chunk)
             LOG.debug(processed_chunk)
-        # Sort found clusters by size, descending order
-        sorted_clusters = sorted(
-            self.miner.drain.clusters, key=lambda it: it.size, reverse=True
-        )
+        clusters = list(self.miner.drain.clusters)
         # Second pass, only matching lines with clusters,
         # to recover original text
-        for chunk_start, chunk in get_chunks(log):
+        for chunk_start, chunk in chunks:
             cluster = self.miner.match(chunk, "always")
-            if cluster in sorted_clusters:
+            if cluster in clusters:
                 out.append((chunk_start, chunk))
-                sorted_clusters.remove(cluster)
+                clusters.remove(cluster)
         return out

logdetective/logdetective.py

@@ -14,8 +14,9 @@ from logdetective.utils import (
     format_snippets,
     compute_certainty,
     load_prompts,
+    load_skip_snippet_patterns,
 )
-from logdetective.extractors import LLMExtractor, DrainExtractor
+from logdetective.extractors import DrainExtractor
 
 LOG = logging.getLogger("logdetective")
 
@@ -49,16 +50,16 @@ def setup_args():
         "--summarizer",
         type=str,
         default="drain",
-        help="Choose between LLM and Drain template miner as the log summarizer.\
-                                LLM must be specified as path to a model, URL or local file.",
+        help="DISABLED: LLM summarization option was removed. \
+                Argument is kept for backward compatibility only.",
     )
     parser.add_argument(
         "-N",
         "--n_lines",
         type=int,
-        default=8,
-        help="The number of lines per chunk for LLM analysis.\
-                            This only makes sense when you are summarizing with LLM.",
+        default=None,
+        help="DISABLED: LLM summarization option was removed. \
+                Argument is kept for backward compatibility only.",
     )
     parser.add_argument(
         "-C",
@@ -74,13 +75,19 @@ def setup_args():
         "--prompts",
         type=str,
         default=f"{os.path.dirname(__file__)}/prompts.yml",
-        help="Path to prompt configuration file."
+        help="Path to prompt configuration file.",
     )
     parser.add_argument(
         "--temperature",
         type=float,
         default=DEFAULT_TEMPERATURE,
-        help="Temperature for inference."
+        help="Temperature for inference.",
+    )
+    parser.add_argument(
+        "--skip_snippets",
+        type=str,
+        default=f"{os.path.dirname(__file__)}/skip_snippets.yml",
+        help="Path to patterns for skipping snippets.",
     )
     return parser.parse_args()
 
@@ -93,6 +100,10 @@ async def run():  # pylint: disable=too-many-statements,too-many-locals
         sys.stderr.write("Error: --quiet and --verbose is mutually exclusive.\n")
         sys.exit(2)
 
+    # Emit warning about use of discontinued args
+    if args.n_lines or args.summarizer != "drain":
+        LOG.warning("LLM based summarization was removed. Drain will be used instead.")
+
     # Logging facility setup
     log_level = logging.INFO
     if args.verbose >= 1:
@@ -116,18 +127,19 @@ async def run():  # pylint: disable=too-many-statements,too-many-locals
         LOG.error("You likely do not have enough memory to load the AI model")
         sys.exit(3)
 
-    # Log file summarizer selection and initialization
-    if args.summarizer == "drain":
-        extractor = DrainExtractor(
-            args.verbose > 1, context=True, max_clusters=args.n_clusters
-        )
-    else:
-        summarizer_model = initialize_model(args.summarizer, verbose=args.verbose > 2)
-        extractor = LLMExtractor(
-            summarizer_model,
-            args.verbose > 1,
-            prompts_configuration.summarization_prompt_template,
-        )
+    try:
+        skip_snippets = load_skip_snippet_patterns(args.skip_snippets)
+    except OSError as e:
+        LOG.error(e)
+        sys.exit(5)
+
+    # Log file summarizer initialization
+    extractor = DrainExtractor(
+        args.verbose > 1,
+        context=True,
+        max_clusters=args.n_clusters,
+        skip_snippets=skip_snippets,
+    )
 
     LOG.info("Getting summary")
 
@@ -151,7 +163,8 @@ async def run():  # pylint: disable=too-many-statements,too-many-locals
 
     prompt = (
         f"{prompts_configuration.default_system_prompt}\n"
-        f"{prompts_configuration.prompt_template}")
+        f"{prompts_configuration.prompt_template}"
+    )
 
     stream = True
     if args.no_stream:
@@ -191,7 +204,7 @@ async def run():  # pylint: disable=too-many-statements,too-many-locals
 
 
 def main():
-    """ Evaluate logdetective program and wait for it to finish """
+    """Evaluate logdetective program and wait for it to finish"""
     asyncio.run(run())
 
 

logdetective/models.py

@@ -1,10 +1,10 @@
+import re
 from typing import Optional
-from pydantic import BaseModel
+from pydantic import BaseModel, model_validator
 
 from logdetective.constants import (
     PROMPT_TEMPLATE,
     PROMPT_TEMPLATE_STAGED,
-    SUMMARIZATION_PROMPT_TEMPLATE,
     SNIPPET_PROMPT_TEMPLATE,
     DEFAULT_SYSTEM_PROMPT,
 )
@@ -14,7 +14,6 @@ class PromptConfig(BaseModel):
     """Configuration for basic log detective prompts."""
 
     prompt_template: str = PROMPT_TEMPLATE
-    summarization_prompt_template: str = SUMMARIZATION_PROMPT_TEMPLATE
     snippet_prompt_template: str = SNIPPET_PROMPT_TEMPLATE
     prompt_template_staged: str = PROMPT_TEMPLATE_STAGED
 
@@ -27,9 +26,6 @@ class PromptConfig(BaseModel):
         if data is None:
             return
         self.prompt_template = data.get("prompt_template", PROMPT_TEMPLATE)
-        self.summarization_prompt_template = data.get(
-            "summarization_prompt_template", SUMMARIZATION_PROMPT_TEMPLATE
-        )
         self.snippet_prompt_template = data.get(
             "snippet_prompt_template", SNIPPET_PROMPT_TEMPLATE
         )
@@ -45,3 +41,33 @@ class PromptConfig(BaseModel):
         self.staged_system_prompt = data.get(
             "staged_system_prompt", DEFAULT_SYSTEM_PROMPT
         )
+
+
+class SkipSnippets(BaseModel):
+    """Regular expressions defining snippets we should not analyze"""
+
+    snippet_patterns: dict[str, re.Pattern] = {}
+
+    def __init__(self, data: Optional[dict] = None):
+        super().__init__(data=data)
+        if data is None:
+            return
+        self.snippet_patterns = {
+            key: re.compile(pattern) for key, pattern in data.items()
+        }
+
+    @model_validator(mode="before")
+    @classmethod
+    def check_patterns(cls, data: dict):
+        """Check if all supplied patterns are valid regular expressions.
+        Techically replicating what is done in __init__ but with nicer error message."""
+        patterns = data["data"]
+        for key, pattern in patterns.items():
+            try:
+                re.compile(pattern=pattern)
+            except (TypeError, re.error) as ex:
+                raise ValueError(
+                    f"Invalid pattern `{pattern}` with name `{key}` supplied for skipping in logs."
+                ) from ex
+
+        return data

logdetective/prompts.yml

@@ -21,17 +21,6 @@ prompt_template: |
 
   Analysis:
 
-
-summarization_prompt_template: |
-  Does following log contain error or issue?
-
-  Log:
-
-  {}
-
-  Answer:
-
-
 snippet_prompt_template: |
   Analyse following RPM build log snippet. Describe contents accurately, without speculation or suggestions for resolution.
 

logdetective/remote_log.py

@@ -53,7 +53,7 @@ class RemoteLog:
             LOG.debug("process url %s", self.url)
             try:
                 response = await self._http_session.get(self.url, raise_for_status=True)
-            except aiohttp.ClientResponseError as ex:
+            except (aiohttp.ClientResponseError, aiohttp.ClientConnectorError) as ex:
                 raise RuntimeError(f"We couldn't obtain the logs: {ex}") from ex
             return await response.text()
         LOG.error("Invalid URL received ")
@@ -64,6 +64,4 @@ class RemoteLog:
         try:
             return await self.get_url_content()
         except RuntimeError as ex:
-            raise HTTPBadRequest(
-                reason=f"We couldn't obtain the logs: {ex}"
-            ) from ex
+            raise HTTPBadRequest(reason=f"We couldn't obtain the logs: {ex}") from ex

logdetective/server/config.py

@@ -3,8 +3,9 @@ import logging
 import yaml
 from openai import AsyncOpenAI
 
-from logdetective.utils import load_prompts
+from logdetective.utils import load_prompts, load_skip_snippet_patterns
 from logdetective.server.models import Config, InferenceConfig
+import logdetective
 
 
 def load_server_config(path: str | None) -> Config:
@@ -52,18 +53,24 @@ def get_log(config: Config):
 
 
 def get_openai_api_client(ineference_config: InferenceConfig):
-    """Set up AsyncOpenAI client with default configuration.
-    """
+    """Set up AsyncOpenAI client with default configuration."""
     return AsyncOpenAI(
-        api_key=ineference_config.api_token,
-        base_url=ineference_config.url)
+        api_key=ineference_config.api_token, base_url=ineference_config.url
+    )
 
 
 SERVER_CONFIG_PATH = os.environ.get("LOGDETECTIVE_SERVER_CONF", None)
 SERVER_PROMPT_PATH = os.environ.get("LOGDETECTIVE_PROMPTS", None)
+# The default location for skip patterns is in the same directory
+# as logdetective __init__.py file.
+SERVER_SKIP_PATTERNS_PATH = os.environ.get(
+    "LOGDETECIVE_SKIP_PATTERNS",
+    f"{os.path.dirname(logdetective.__file__)}/skip_snippets.yml",
+)
 
 SERVER_CONFIG = load_server_config(SERVER_CONFIG_PATH)
 PROMPT_CONFIG = load_prompts(SERVER_PROMPT_PATH)
+SKIP_SNIPPETS_CONFIG = load_skip_snippet_patterns(SERVER_SKIP_PATTERNS_PATH)
 
 LOG = get_log(SERVER_CONFIG)
 

logdetective/server/emoji.py

@@ -51,7 +51,9 @@ async def _handle_gitlab_operation(func: Callable, *args):
         else:
             LOG.exception(log_msg)
     except Exception as e:  # pylint: disable=broad-exception-caught
-        LOG.exception("Unexpected error during GitLab operation %s(%s): %s", func, args, e)
+        LOG.exception(
+            "Unexpected error during GitLab operation %s(%s): %s", func, args, e
+        )
 
 
 async def collect_emojis_in_comments(  # pylint: disable=too-many-locals

logdetective/server/llm.py

@@ -16,7 +16,13 @@ from logdetective.utils import (
     compute_certainty,
     prompt_to_messages,
 )
-from logdetective.server.config import LOG, SERVER_CONFIG, PROMPT_CONFIG, CLIENT
+from logdetective.server.config import (
+    LOG,
+    SERVER_CONFIG,
+    PROMPT_CONFIG,
+    CLIENT,
+    SKIP_SNIPPETS_CONFIG,
+)
 from logdetective.server.models import (
     AnalyzedSnippet,
     InferenceConfig,
@@ -42,7 +48,10 @@ def format_analyzed_snippets(snippets: list[AnalyzedSnippet]) -> str:
 def mine_logs(log: str) -> List[Tuple[int, str]]:
     """Extract snippets from log text"""
     extractor = DrainExtractor(
-        verbose=True, context=True, max_clusters=SERVER_CONFIG.extractor.max_clusters
+        verbose=True,
+        context=True,
+        max_clusters=SERVER_CONFIG.extractor.max_clusters,
+        skip_snippets=SKIP_SNIPPETS_CONFIG,
     )
 
     LOG.info("Getting summary")

logdetective/server/plot.py

@@ -2,12 +2,10 @@ import datetime
 from typing import Optional, Union, Dict
 
 import numpy
-import matplotlib
-import matplotlib.figure
-import matplotlib.pyplot
+from numpy.typing import ArrayLike
+from matplotlib import dates, colormaps, axes, pyplot, figure
 
-from matplotlib.pyplot import cm
-from logdetective.server import models
+from logdetective.server.models import TimePeriod
 from logdetective.server.database.models import (
     AnalyzeRequestMetrics,
     EndpointType,
@@ -18,25 +16,25 @@ from logdetective.server.database.models import (
 class Definition:
     """Define plot details, given a time period."""
 
-    def __init__(self, time_period: models.TimePeriod):
+    def __init__(self, time_period: TimePeriod):
         self.time_period = time_period
         self.days_diff = time_period.get_time_period().days
         if self.time_period.hours:
             self._freq = "H"
             self._time_format = "%Y-%m-%d %H"
-            self._locator = matplotlib.dates.HourLocator(interval=2)
+            self._locator = dates.HourLocator(interval=2)
             self._time_unit = "hour"
             self._time_delta = datetime.timedelta(hours=1)
         elif self.time_period.days:
             self._freq = "D"
             self._time_format = "%Y-%m-%d"
-            self._locator = matplotlib.dates.DayLocator(interval=1)
+            self._locator = dates.DayLocator(interval=1)
             self._time_unit = "day"
             self._time_delta = datetime.timedelta(days=1)
         elif self.time_period.weeks:
             self._freq = "W"
             self._time_format = "%Y-%m-%d"
-            self._locator = matplotlib.dates.WeekdayLocator(interval=1)
+            self._locator = dates.WeekdayLocator(interval=1)
             self._time_unit = "week"
             self._time_delta = datetime.timedelta(weeks=1)
 
@@ -120,10 +118,10 @@ def create_time_series_arrays(
 
 
 def _add_bar_chart(
-    ax: matplotlib.figure.Axes,
+    ax: axes.Axes,
     plot_def: Definition,
-    timestamps: numpy.array,
-    values: numpy.array,
+    timestamps: ArrayLike,
+    values: ArrayLike,
     label: str,
 ) -> None:
     """Add a blue bar chart"""
@@ -142,18 +140,18 @@ def _add_bar_chart(
     ax.set_ylabel(label, color="blue")
     ax.tick_params(axis="y", labelcolor="blue")
 
-    ax.xaxis.set_major_formatter(matplotlib.dates.DateFormatter(plot_def.time_format))
+    ax.xaxis.set_major_formatter(dates.DateFormatter(plot_def.time_format))
     ax.xaxis.set_major_locator(plot_def.locator)
 
-    matplotlib.pyplot.xticks(rotation=45)
+    pyplot.xticks(rotation=45)
 
     ax.grid(True, alpha=0.3)
 
 
 def _add_line_chart(  # pylint: disable=too-many-arguments disable=too-many-positional-arguments
-    ax: matplotlib.figure.Axes,
-    timestamps: numpy.array,
-    values: numpy.array,
+    ax: axes.Axes,
+    timestamps: ArrayLike,
+    values: ArrayLike,
     label: str,
     color: str = "red",
     set_label: bool = True,
@@ -166,10 +164,10 @@ def _add_line_chart(  # pylint: disable=too-many-arguments disable=too-many-posi
 
 
 def requests_per_time(
-    period_of_time: models.TimePeriod,
+    period_of_time: TimePeriod,
     endpoint: EndpointType = EndpointType.ANALYZE,
     end_time: Optional[datetime.datetime] = None,
-) -> matplotlib.figure.Figure:
+) -> figure.Figure:
     """
     Generate a visualization of request counts over a specified time period.
 
@@ -200,13 +198,13 @@ def requests_per_time(
         requests_counts, plot_def, start_time, end_time
     )
 
-    fig, ax1 = matplotlib.pyplot.subplots(figsize=(12, 6))
+    fig, ax1 = pyplot.subplots(figsize=(12, 6))
     _add_bar_chart(ax1, plot_def, timestamps, counts, "Requests")
 
     ax2 = ax1.twinx()
     _add_line_chart(ax2, timestamps, numpy.cumsum(counts), "Cumulative Requests")
 
-    matplotlib.pyplot.title(
+    pyplot.title(
         f"Requests received for API {endpoint} ({start_time.strftime(plot_def.time_format)} "
         f"to {end_time.strftime(plot_def.time_format)})"
     )
@@ -215,16 +213,16 @@ def requests_per_time(
     lines2, labels2 = ax2.get_legend_handles_labels()
     ax1.legend(lines1 + lines2, labels1 + labels2, loc="center")
 
-    matplotlib.pyplot.tight_layout()
+    pyplot.tight_layout()
 
     return fig
 
 
 def average_time_per_responses(  # pylint: disable=too-many-locals
-    period_of_time: models.TimePeriod,
+    period_of_time: TimePeriod,
     endpoint: EndpointType = EndpointType.ANALYZE,
     end_time: Optional[datetime.datetime] = None,
-) -> matplotlib.figure.Figure:
+) -> figure.Figure:
     """
     Generate a visualization of average response time and length over a specified time period.
 
@@ -259,7 +257,7 @@ def average_time_per_responses(  # pylint: disable=too-many-locals
         float,
     )
 
-    fig, ax1 = matplotlib.pyplot.subplots(figsize=(12, 6))
+    fig, ax1 = pyplot.subplots(figsize=(12, 6))
     _add_bar_chart(
         ax1, plot_def, timestamps, average_time, "average response time (seconds)"
     )
@@ -280,7 +278,7 @@ def average_time_per_responses(  # pylint: disable=too-many-locals
     ax2 = ax1.twinx()
     _add_line_chart(ax2, timestamps, average_length, "average response length (chars)")
 
-    matplotlib.pyplot.title(
+    pyplot.title(
         f"average response time for API {endpoint} ({start_time.strftime(plot_def.time_format)} "
         f"to {end_time.strftime(plot_def.time_format)})"
     )
@@ -289,7 +287,7 @@ def average_time_per_responses(  # pylint: disable=too-many-locals
     lines2, labels2 = ax2.get_legend_handles_labels()
     ax1.legend(lines1 + lines2, labels1 + labels2, loc="center")
 
-    matplotlib.pyplot.tight_layout()
+    pyplot.tight_layout()
 
     return fig
 
@@ -322,7 +320,7 @@ def _collect_emoji_data(
 
 
 def _plot_emoji_data(  # pylint: disable=too-many-locals
-    ax: matplotlib.figure.Axes,
+    ax: axes.Axes,
     reactions_values_dict: Dict[str, Dict[datetime.datetime, int]],
     plot_def: Definition,
     start_time: datetime.datetime,
@@ -340,7 +338,10 @@ def _plot_emoji_data(  # pylint: disable=too-many-locals
         )
         all_counts.extend(counts)
 
-    colors = [cm.viridis(i) for i in numpy.linspace(0, 1, len(reactions_values_dict))]    # pylint: disable=no-member
+    colors = [
+        colormaps["viridis"](i)
+        for i in numpy.linspace(0, 1, len(reactions_values_dict))
+    ]
 
     first_emoji = True
     for i, (emoji, dict_counts) in enumerate(reactions_values_dict.items()):
@@ -369,9 +370,9 @@ def _plot_emoji_data(  # pylint: disable=too-many-locals
 
 
 def emojis_per_time(
-    period_of_time: models.TimePeriod,
+    period_of_time: TimePeriod,
     end_time: Optional[datetime.datetime] = None,
-) -> matplotlib.figure.Figure:
+) -> figure.Figure:
     """
     Generate a visualization of overall emoji feedback
     over a specified time period.
@@ -396,13 +397,13 @@ def emojis_per_time(
     start_time = period_of_time.get_period_start_time(end_time)
     reactions_values_dict = _collect_emoji_data(start_time, plot_def)
 
-    fig, ax = matplotlib.pyplot.subplots(figsize=(12, 6))
+    fig, ax = pyplot.subplots(figsize=(12, 6))
 
     emoji_lines, emoji_labels = _plot_emoji_data(
         ax, reactions_values_dict, plot_def, start_time, end_time
     )
 
-    matplotlib.pyplot.title(
+    pyplot.title(
         f"Emoji feedback ({start_time.strftime(plot_def.time_format)} "
         f"to {end_time.strftime(plot_def.time_format)})"
     )
@@ -419,11 +420,11 @@ def emojis_per_time(
     ax.set_ylabel("Count")
 
     # Format x-axis
-    ax.xaxis.set_major_formatter(matplotlib.dates.DateFormatter(plot_def.time_format))
+    ax.xaxis.set_major_formatter(dates.DateFormatter(plot_def.time_format))
     ax.xaxis.set_major_locator(plot_def.locator)
     ax.tick_params(axis="x", labelrotation=45)
     ax.grid(True, alpha=0.3)
 
-    matplotlib.pyplot.tight_layout()
+    pyplot.tight_layout()
 
     return fig

logdetective/skip_snippets.yml

@@ -0,0 +1,12 @@
+# This file holds patterns you want to skip during log parsing.
+# By default, no patterns are supplied.
+# Patterns are to be specified as values of dictionary,
+# with each key being a descriptive name of the pattern.
+# Patterns themselves are evaluated as a regular expression.
+# Make sure to avoid regular expressions that may be interpreted
+# as yaml syntax.
+# Example:
+
+# contains_capital_a: "^.*A.*"
+# starts_with_numeric: "^[0-9].*"
+child_exit_code_zero: "Child return code was: 0"

logdetective/utils.py

@@ -8,7 +8,7 @@ import numpy as np
 import yaml
 
 from llama_cpp import Llama, CreateCompletionResponse, CreateCompletionStreamResponse
-from logdetective.models import PromptConfig
+from logdetective.models import PromptConfig, SkipSnippets
 from logdetective.remote_log import RemoteLog
 
 
@@ -179,7 +179,7 @@ def format_snippets(snippets: list[str] | list[Tuple[int, str]]) -> str:
             summary += f"""
             Snippet No. {i}:
 
-            {s[1]}
+            {s}
             ================
             """
     return summary
@@ -198,8 +198,11 @@ def load_prompts(path: str | None) -> PromptConfig:
 
 
 def prompt_to_messages(
-        user_message: str, system_prompt: str | None = None,
-        system_role: str = "developer", user_role: str = "user") -> List[Dict[str, str]]:
+    user_message: str,
+    system_prompt: str | None = None,
+    system_role: str = "developer",
+    user_role: str = "user",
+) -> List[Dict[str, str]]:
     """Turn prompt into list of message dictionaries.
     If `system_role` and `user_role` are the same, only a single message is created,
     as concatenation of `user_message` and `system_prompt`. This is useful for models which
@@ -208,22 +211,39 @@ def prompt_to_messages(
 
     if system_role == user_role:
         messages = [
-            {
-                "role": system_role,
-                "content": f"{system_prompt}\n{user_message}"
-            }
+            {"role": system_role, "content": f"{system_prompt}\n{user_message}"}
         ]
     else:
-
         messages = [
-            {
-                "role": system_role,
-                "content": system_prompt
-            },
+            {"role": system_role, "content": system_prompt},
             {
                 "role": user_role,
                 "content": user_message,
-            }
+            },
         ]
 
     return messages
+
+
+def filter_snippet_patterns(snippet: str, skip_snippets: SkipSnippets) -> bool:
+    """Try to match snippet agains provided patterns to determine if we should
+    filter it out or not."""
+    for key, pattern in skip_snippets.snippet_patterns.items():
+        if pattern.match(snippet):
+            LOG.debug("Snippet `%s` has matched agains skip pattern %s", snippet, key)
+            return True
+
+    return False
+
+
+def load_skip_snippet_patterns(path: str | None) -> SkipSnippets:
+    """Load dictionary of snippet patterns we want to skip."""
+    if path:
+        try:
+            with open(path, "r") as file:
+                return SkipSnippets(yaml.safe_load(file))
+        except OSError as e:
+            LOG.error("Couldn't open file with snippet skip patterns `%s`", path)
+            raise e
+
+    return SkipSnippets({})

{logdetective-1.4.0.dist-info → logdetective-1.6.0.dist-info}/METADATA

@@ -1,12 +1,12 @@
 Metadata-Version: 2.3
 Name: logdetective
-Version: 1.4.0
+Version: 1.6.0
 Summary: Log using LLM AI to search for build/test failures and provide ideas for fixing these.
 License: Apache-2.0
 Author: Jiri Podivin
 Author-email: jpodivin@gmail.com
 Requires-Python: >=3.11,<4.0
-Classifier: Development Status :: 4 - Beta
+Classifier: Development Status :: 5 - Production/Stable
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: Apache Software License
@@ -87,9 +87,10 @@ Usage
 To analyze a log file, run the script with the following command line arguments:
 - `url` (required): The URL of the log file to be analyzed.
 - `--model` (optional, default: "Mistral-7B-Instruct-v0.2-GGUF"): The path or URL of the language model for analysis. As we are using LLama.cpp we want this to be in the `gguf` format. You can include the download link to the model here. If the model is already on your machine it will skip the download.
-- `--summarizer` (optional, default: "drain"): Choose between LLM and Drain template miner as the log summarizer. You can also provide the path to an existing language model file instead of using a URL.
-- `--n_lines` (optional, default: 8): The number of lines per chunk for LLM analysis. This only makes sense when you are summarizing with LLM.
+- `--summarizer` DISABLED: LLM summarization option was removed. Argument is kept for backward compatibility only.(optional, default: "drain"): Choose between LLM and Drain template miner as the log summarizer. You can also provide the path to an existing language model file instead of using a URL.
+- `--n_lines` DISABLED: LLM summarization option was removed. Argument is kept for backward compatibility only. (optional, default: 8): The number of lines per chunk for LLM analysis. This only makes sense when you are summarizing with LLM.
 - `--n_clusters` (optional, default 8): Number of clusters for Drain to organize log chunks into. This only makes sense when you are summarizing with Drain
+- `--skip_snippets` Path to patterns for skipping snippets.
 
 Example usage:
 
@@ -376,6 +377,9 @@ HTTPS certificate generated through:
 certbot certonly --standalone -d logdetective01.fedorainfracloud.org
 ```
 
+Certificates need to be be placed into location specified by the`LOGDETECTIVE_CERTDIR`
+env var and the service should be restarted.
+
 Querying statistics
 -------------------
 
@@ -435,6 +439,28 @@ with spaces, or replacement fields marked with curly braces, `{}` left for inser
 Number of replacement fields in new prompts, must be the same as in originals.
 Although their position may be different.
 
+
+Skip Snippets
+-------------
+
+Certain log chunks may not contribute to the analysis of the problem under any circumstances.
+User can specify regular expressions, matching such log chunks, along with simple description,
+using Skip Snippets feature.
+
+Patterns to be skipped must be defined yaml file as a dictionary, where key is a description
+and value is a regular expression. For example:
+
+```
+child_exit_code_zero: "Child return code was: 0"
+```
+
+Special care must be taken not to write a regular expression which may match
+too many chunks, or which may be evaluated as data structure by the yaml parser.
+
+Example of a valid pattern definition file: `logdetective/skip_patterns.yml`,
+can be used as a starting point and is used as a default if no other definition is provided.
+
+
 License
 -------
 

{logdetective-1.4.0.dist-info → logdetective-1.6.0.dist-info}/RECORD

@@ -1,33 +1,34 @@
 logdetective/__init__.py,sha256=VqRngDcuFT7JWms8Qc_MsOvajoXVOKPr-S1kqY3Pqhc,59
-logdetective/constants.py,sha256=yH2vpYENfJpoYUC6KzzTSRDFOiMmjtRK-GU9vfv8o3o,2560
+logdetective/constants.py,sha256=KD5FtMvRMO5jO9O1a5FbHy6yFSF6ZkZ4lNrhI7D_S2Y,2456
 logdetective/drain3.ini,sha256=ni91eCT1TwTznZwcqWoOVMQcGEnWhEDNCoTPF7cfGfY,1360
-logdetective/extractors.py,sha256=7ahzWbTtU9MveG1Q7wU9LO8OJgs85X-cHmWltUhCe9M,3491
-logdetective/logdetective.py,sha256=YwjKSgW2iW4BAxWTOgUbrVbk2Lbv76vOLFTTtuPg3hg,5928
-logdetective/models.py,sha256=0DHCtGUqZzPcduSU4Z5AzuNn3g9XBh6UhLBcA9mDG6M,1653
+logdetective/extractors.py,sha256=BkQe7FMLDoKVWitP85Vpv1qle1Fo1FeupKm0wVlcALI,1859
+logdetective/logdetective.py,sha256=Wx3osTvVjrt8EqpXsq7YbqIOJ_z_7p3k8HN2QURQ2hQ,6153
+logdetective/models.py,sha256=h01nagxgb8sR9Js_00DMoZv6GvwHjcOk0MeKttftDHk,2460
 logdetective/prompts-summary-first.yml,sha256=3Zfp4NNOfaFYq5xBlBjeQa5PdjYfS4v17OtJqQ-DRpU,821
 logdetective/prompts-summary-only.yml,sha256=8U9AMJV8ePW-0CoXOXlQoO92DAJDeutIT8ntSkkm6W0,470
-logdetective/prompts.yml,sha256=2WGox--VRINjsxaJc7fql-v1sIVYWYZLEjJV_vvBm9o,3969
-logdetective/remote_log.py,sha256=1oeMIdDE_ob_2QrlXYTAA_m-36pNEicXbZwrCyzNgwo,2256
+logdetective/prompts.yml,sha256=dOqaFrtBOkFRHLWte_2tGV-pNXwXP9Km9iWno_TZyic,3863
+logdetective/remote_log.py,sha256=28QvdQiy7RBnd86EKCq_A75P21gSNlCbgxJe5XAe9MA,2258
 logdetective/server/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 logdetective/server/compressors.py,sha256=qzrT-BPSksXY6F2L6ger04GGrgdBsGOfK2YuCFRs0Q4,5427
-logdetective/server/config.py,sha256=yD6pRTZze9bqoCxfdxpupXrvb18auAf95mq4BoowAbs,2113
+logdetective/server/config.py,sha256=Qg3Q15S-NPXEpSqGgAH41JVy4dheqKYUSEkzE-1Z05E,2478
 logdetective/server/database/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 logdetective/server/database/base.py,sha256=1mcjEbhwLl4RalvT3oy6XVctjJoWIW3H9aI_sMWJBK8,1728
 logdetective/server/database/models/__init__.py,sha256=xy2hkygyw6_87zPKkG20i7g7_LXTGR__PUeojhbvv94,496
 logdetective/server/database/models/merge_request_jobs.py,sha256=hw88wV1-3x7i53sX7ZotKClc6OsH1njPpbRSZofnqr4,18670
 logdetective/server/database/models/metrics.py,sha256=yl9fS4IPVFWDeFvPAxO6zOVu6oLF319ApvVLAgnD5yU,13928
-logdetective/server/emoji.py,sha256=W1nJLU1UnTG8FGttOs6gC7x3TcjxiBuviXklD9f2Mu8,4370
+logdetective/server/emoji.py,sha256=Iv1CFNyWhweBG13v59O1fQD-dZj-YGM1IKlkIaCzBaU,4392
 logdetective/server/gitlab.py,sha256=wQSlvdWn6XEi1oP6HhI75bIhm6bgdpWr3zu2WXF0_oE,16473
-logdetective/server/llm.py,sha256=q9LdoAmsx9MpBjnjLyJ9GBU27jKViTaWbVXyMsmsCI0,5721
+logdetective/server/llm.py,sha256=EiLp3QV3OAvZcqrq6t21M0vzHFiPuVMamRLWPggqTEo,5829
 logdetective/server/metric.py,sha256=B3ew_qSmtEMj6xl-FoOtS4F_bkplp-shhtfHF1cG_Io,4010
 logdetective/server/models.py,sha256=I45uLnq_zqn_r0FdOdop9zQPbsOWOY_M39NBBOXP134,15738
-logdetective/server/plot.py,sha256=eZs4r9gua-nW3yymSMIz1leL9mb4QKlh6FJZSeOfZ5M,14872
+logdetective/server/plot.py,sha256=C98U9prGoPkp8_t4v2dovdZuwOhSbxXSeB_K9Q2r3NE,14607
 logdetective/server/server.py,sha256=V-lSG2cCTxoGwvUc8mEmLQQWS4g_W_dER2o118RufAk,18792
 logdetective/server/templates/gitlab_full_comment.md.j2,sha256=DQZ2WVFedpuXI6znbHIW4wpF9BmFS8FaUkowh8AnGhE,1627
 logdetective/server/templates/gitlab_short_comment.md.j2,sha256=fzScpayv2vpRLczP_0O0YxtA8rsKvR6gSv4ntNdWb98,1443
-logdetective/utils.py,sha256=4VDghJs6mTz8PjkYaV794LUWWyEqBRddXKkENJzR5n4,7025
-logdetective-1.4.0.dist-info/LICENSE,sha256=z8d0m5b2O9McPEK1xHG_dWgUBT6EfBDz6wA0F7xSPTA,11358
-logdetective-1.4.0.dist-info/METADATA,sha256=UlBDdFwH4kswo5NqhvEWfz7dGcex7aXfFOUt1JI5n4Y,17709
-logdetective-1.4.0.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
-logdetective-1.4.0.dist-info/entry_points.txt,sha256=3K_vXja6PmcA8sNdUi63WdImeiNhVZcEGPTaoJmltfA,63
-logdetective-1.4.0.dist-info/RECORD,,
+logdetective/skip_snippets.yml,sha256=reGlhPPCo06nNUJWiC2LY-OJOoPdcyOB7QBTSMeh0eg,487
+logdetective/utils.py,sha256=UAhPWbOGdTR7PWc1dEQk8FSxBSsO0UgfdyY8AKGfMJY,7781
+logdetective-1.6.0.dist-info/LICENSE,sha256=z8d0m5b2O9McPEK1xHG_dWgUBT6EfBDz6wA0F7xSPTA,11358
+logdetective-1.6.0.dist-info/METADATA,sha256=PbYrNQMqhQ257mdwsSjvlDBX6abVFsY-Ky9P5WsBtv8,18891
+logdetective-1.6.0.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+logdetective-1.6.0.dist-info/entry_points.txt,sha256=3K_vXja6PmcA8sNdUi63WdImeiNhVZcEGPTaoJmltfA,63
+logdetective-1.6.0.dist-info/RECORD,,