Merge branch 'dthomas_meilisearch'

# Conflicts:
#	.gitignore
#	read_datasheet.py

.gitignore

@@ -1,10 +1,18 @@
+# python
 venv
 __pycache__
+# cable data folder(s)
 cables
+cables-sample.zip
+# meilisearch (mainly where I've put the data volume for the container)
+meili_data
+# IDE things
 .vscode
-output.log
+.idea
+# videos
 *.webm
 output.mp4
+# log files
 output.log
-cables-sample.zip
-*.png
+# images
+*.png

compose.yml

@@ -0,0 +1,13 @@
+services:
+  meilisearch:
+    image: "getmeili/meilisearch:v1.6.2"
+    ports:
+      - "7700:7700"
+    environment:
+      MEILI_MASTER_KEY: fluffybunnyrabbit
+      MEILI_NO_ANALYTICS: true
+    volumes:
+      - "meili_data:/meili_data"
+
+volumes:
+  meili_data:

database.py

@@ -1,140 +0,0 @@
-"""This module contains functionality for interacting with a PostgreSQL database. It will automatically handle error
-conditions (i.e. missing columns) without terminating the entire program. Use the :py:class:`DBConnector` class to
-handle database interactions, either as a standalone object or in a context manager."""
-from __future__ import annotations
-
-import os
-import psycopg2
-from psycopg2 import DatabaseError, OperationalError
-from psycopg2.errors import UndefinedColumn
-
-DB_ADDRESS = os.getenv('DB_ADDRESS', 'localhost')
-DB_PORT = os.getenv('DB_PORT', 5432)
-DB_USER = os.getenv('DB_USER', 'postgres')
-DB_PASSWORD = os.getenv('DB_PASSWORD', '')
-DB_NAME = os.getenv('DB_NAME', 'postgres')
-DB_TABLE = os.getenv('DB_TABLE', 'cables')
-
-
-class DBConnector:
-    """Context managed database class. Use with statements to automatically open and close the database connection, like
-    so:
-
-    .. code-block:: python
-       with DBConnector() as db:
-           db.read()
-    """
-
-    def _db_start(self):
-        """Setup the database connection and cursor."""
-        try:
-            self.conn = psycopg2.connect(
-                f"host={DB_ADDRESS} port={DB_PORT} dbname={DB_NAME} user={DB_USER} password={DB_PASSWORD}")
-            self.cur = self.conn.cursor()
-        except OperationalError as e:
-            raise e
-
-    def _db_stop(self):
-        """Close the cursor and connection."""
-        self.cur.close()
-        self.conn.close()
-
-    def __init__(self):
-        self._db_start()
-
-    def __del__(self):
-        self._db_stop()
-
-    def __enter__(self):
-        self._db_start()
-
-    def __exit__(self):
-        self._db_stop()
-
-    def _get_cols(self) -> set[str]:
-        """Get the list of columns in the database.
-
-        :return: A list of column names."""
-        query = f"select COLUMN_NAME from information_schema.columns where table_name={DB_TABLE}"
-        rows = {x["COLUMN_NAME"] for x in self._query(query)}
-        return rows
-
-    def _column_parity(self, columns: list[str] | set[str]) -> set[str]:
-        """If the listed columns are not in the database, add them.
-
-        :param columns: The columns we expect are in the database.
-        :return: The list of columns in the database after querying."""
-        cols = set(columns)
-        existing = self._get_cols()
-        needs = cols.difference(existing.intersection(cols))
-        if len(needs) > 0:
-            query = f"ALTER TABLE {DB_TABLE} {', '.join([f'ADD COLUMN {c}' for c in needs])}"
-            self._query(query)
-            existing = self._get_cols()
-        return existing
-
-    def _query(self, sql) -> list[dict]:
-        """Basic function for running queries.
-
-        :param sql: SQL query as plaintext.
-        :return: Results of the query, or an empty list if none."""
-        result = []
-        try:
-            self.cur.execute(sql)
-            result = self._read_dict()
-        except DatabaseError as e:
-            print(f"ERROR {e.pgcode}: {e.pgerror}\n"
-                  f"Caused by query: {sql}")
-        finally:
-            return result
-
-    def _read_dict(self) -> list[dict]:
-        """Read the cursor as a list of dictionaries. psycopg2 defaults to using a list of tuples, so we want to convert
-        each row into a dictionary before we return it."""
-        cols = [i.name for i in self.cur.description]
-        results = []
-        for row in self.cur:
-            row_dict = {}
-            for i in range(0, len(row)):
-                if row[i]:
-                    row_dict = {**row_dict, cols[i]: row[i]}
-            results.append(row_dict)
-        return results
-
-    def read(self, **kwargs) -> list[dict]:
-        """Read rows from a database that match the specified filters.
-
-        :param kwargs: Column constraints; i.e. what value to filter by in what column.
-        :returns: A list of dictionaries of all matching rows, or an empty list if no match."""
-        args = []
-        for kw in kwargs.keys():
-            args.append(f"{kw} ILIKE {kwargs['kw']}")
-        query = f"SELECT * FROM {DB_TABLE}"
-        if len(args) > 0:
-            query += f" WHERE {' AND '.join(args)}"
-        return self._query(query)
-
-    def write(self, **kwargs) -> dict:
-        """Write a row to the database.
-
-        :param kwargs: Values to write for each database; specify each column separately!
-        :returns: The row you just added."""
-        self._column_parity(set(kwargs.keys()))
-        values = []
-        for val in kwargs.keys():
-            values.append(kwargs[val])
-        query = f"INSERT INTO {DB_TABLE} ({', '.join(kwargs.keys())}) VALUES ({', '.join(values)})"
-        self._query(query)
-        return kwargs
-
-    def write_all(self, items: list[dict]) -> list[dict]:
-        """Write multiple rows to the database.
-
-        :param items: Rows to write, as a list of dictionaries.
-        :returns: The rows that were added successfully."""
-        successes = []
-        for i in items:
-            res0 = self.write(**i)
-            if res0:
-                successes.append(res0)
-        return successes

read_datasheet.py

@@ -30,7 +30,7 @@ def parse(filename, output_dir, partnum, dstype):
     reader = PdfReader(filename)
     page = reader.pages[0]
     table_list = {}
-    
+
     for table in tables:
         table.df.infer_objects(copy=False)
         table.df.replace('', np.nan, inplace=True)
@@ -253,6 +253,11 @@ def flatten(tables):
                 out[fullkeyname] = list(map(lambda x: x.strip(), tables[table][key].split(",")))
                 #print("\"" + keyname + "\":", "\"" + str(out[fullkeyname]) + "\",")
 
+            # if the item has at least two commas in it, split it
+            if tables[table][key].count(',') > 0:
+                out[fullkeyname] = list(map(lambda x: x.strip(), tables[table][key].split(",")))
+                print("\"" + keyname + "\":", "\"" + str(out[fullkeyname]) + "\",")
+
 
     #print("}")
     return out

requirements.txt

@@ -5,7 +5,7 @@ pypdf2==2.12.1
 alive-progress
 requests
 git+https://github.com/Byeongdulee/python-urx.git
-psycopg2-binary
+meilisearch
 pyyaml
 Flask
 selenium

search.py

@@ -0,0 +1,117 @@
+"""Interactions with the Meilisearch API for adding and searching cables."""
+from meilisearch import Client
+from meilisearch.task import TaskInfo
+from meilisearch.errors import MeilisearchApiError
+import json
+
+DEFAULT_URL = "http://localhost:7700"
+DEFAULT_APIKEY = "fluffybunnyrabbit" # I WOULD RECOMMEND SOMETHING MORE SECURE
+DEFAULT_INDEX = "cables"
+DEFAULT_FILTERABLE_ATTRS = ["partnum", "uuid", "position"] # default filterable attributes
+
+
+class JukeboxSearch:
+    """Class for interacting with the Meilisearch API."""
+    def __init__(self,
+                 url: str = None,
+                 api_key: str = None,
+                 index: str = None,
+                 filterable_attrs: list = None):
+        """Connect to Meilisearch and perform first-run tasks as necessary.
+
+        :param url: Address of the Meilisearch server. Defaults to ``http://localhost:7700`` if unspecified.
+        :param api_key: API key used to authenticate with Meilisearch. It is highly recommended to set this as something
+        secure if you can access this endpoint publicly, but you can ignore this and set Meilisearch's default API key
+        to ``fluffybunnyrabbit``.
+        :param index: The name of the index to configure. Defaults to ``cables`` if unspecified.
+        :param filterable_attrs: List of all the attributes we want to filter by."""
+        # connect to Meilisearch
+        url = url or DEFAULT_URL
+        api_key = api_key or DEFAULT_APIKEY
+        filterable_attrs = filterable_attrs or DEFAULT_FILTERABLE_ATTRS
+        self.index = index or DEFAULT_INDEX
+        self.client = Client(url, api_key)
+        # create the index if it does not exist already
+        try:
+            self.client.get_index(self.index)
+        except MeilisearchApiError as _:
+            self.client.create_index(self.index)
+        # make a variable to easily reference the index
+        self.idxref = self.client.index(self.index)
+
+        # update filterable attributes if needed
+        self.update_filterables(filterable_attrs)
+
+    def add_document(self, document: dict) -> TaskInfo:
+        """Add a cable to the Meilisearch index.
+
+        :param document: Dictionary containing all the cable data.
+        :returns: A TaskInfo object for the addition of the new document."""
+        return self.idxref.add_documents(document)
+
+    def add_documents(self, documents: list):
+        """Add a list of cables to the Meilisearch index.
+
+        :param documents: List of dictionaries containing all the cable data.
+        :returns: A TaskInfo object for the last new document."""
+        taskinfo = None
+        for i in documents:
+            taskinfo = self.add_document(i)
+        return taskinfo
+    
+    def update_filterables(self, filterables: list):
+        """Update filterable attributes and wait for database to fully index. If the filterable attributes matches the
+        current attributes in the database, don't update (saves reindexing).
+        
+        :param filterables: List of all filterable attributes"""
+
+        existing_filterables = self.idxref.get_filterable_attributes()
+        if len(set(existing_filterables).difference(set(filterables))) > 0:
+            taskref = self.idxref.update_filterable_attributes(filterables)
+
+            self.client.wait_for_task(taskref.index_uid)
+
+    def search(self, query: str, filters: str = None):
+        """Execute a search query on the Meilisearch index.
+        
+        :param query: Seach query
+        :param filters: A meilisearch compatible filter statement.
+        :returns: The search results dict. Actual results are in a list under "hits", but there are other nice values that are useful in the root element."""
+        if filters:
+            q = self.idxref.search(query, {"filter": filters})
+        else:
+            q = self.idxref.search(query)
+        return q
+    
+    def _filter_one(self, filter: str):
+        """Get the first item to match a filter.
+        
+        :param filter: A meilisearch compatible filter statement.
+        :returns: A dict containing the results; If no results found, an empty dict."""
+        q = self.search("", filter)
+        if q["estimatedTotalHits"] != 0:
+            return ["hits"][0]
+        else:
+            return dict()
+
+    def get_position(self, position: str):
+        """Get a part by position.
+        
+        :param partnum: The position to search for."""
+        return self._filter_one(f"position = {position}")
+    
+    def get_uuid(self, uuid: str):
+        """Get a specific UUID.
+        
+        :param uuid: The UUID to search for."""
+        return self._filter_one(f"uuid = {uuid}")
+
+    def get_partnum(self, partnum: str):
+        """Get a specific part number.
+        
+        :param partnum: The part number to search for."""
+        return self._filter_one(f"partnum = {partnum}")
+
+# entrypoint
+if __name__ == "__main__":
+    jbs = JukeboxSearch()