Issue #39 feature/fixes

mcioffi · mcioffi · commit 5c4e541d5492 · 2025-09-02T07:54:58.000-04:00
For #39 - [x] Labeller: Adds `==` operator for ID queries - [x] Labeller: Fixes offsetPage default value (should be 0)
diff --git a/webtools/app.py b/webtools/app.py
@@ -33,6 +33,7 @@
 SQL3_DATABASE_TABLE = "en"
 SQL3_DATABASE = parent_dir / "train/data/training.sqlite3"
 MODEL_REQUIREMENTS = parent_dir / "requirements-dev.txt"
+RESERVED_LABELLER_SEARCH_CHARS = r"\*\*|\~\~|\=\="
 
 # sqlite
 sqlite3.register_adapter(list, json.dumps)
@@ -50,19 +51,19 @@ def error_response(status: int, message: str = ""):
     """Boilerplate for errors"""
     if status == 400:
         return jsonify(
-            {"status": 400, "error": "Sorry, bad params", "message": None}
+            {"status": 400, "error": "Sorry, bad params", "message": message}
         ), 400
     elif status == 404:
         return jsonify(
-            {"status": 404, "error": "Sorry, resource not found", "message": None}
+            {"status": 404, "error": "Sorry, resource not found", "message": message}
         ), 404
     elif status == 500:
         return jsonify(
             {"status": 404, "error": "Sorry, api failed", "message": message}
         ), 500
     else:
         return jsonify(
-            {"status": status, "error": "Sorry, something failed", "message": None}
+            {"status": status, "error": "Sorry, something failed", "message": message}
         ), 500
 
 
@@ -111,13 +112,13 @@ def parser():
 
         try:
             sentence = data.get("sentence", "")
-            discard_isolated_stop_words = data.get("discard_isolated_stop_words", False)
-            expect_name_in_output = data.get("expect_name_in_output", False)
+            discard_isolated_stop_words = data.get("discard_isolated_stop_words", True)
+            expect_name_in_output = data.get("expect_name_in_output", True)
             string_units = data.get("string_units", False)
             imperial_units = data.get("imperial_units", False)
-            foundation_foods = data.get("foundation_foods", False)
+            foundation_foods = data.get("foundation_foods", True)
             optimistic_cache_reset = data.get("optimistic_cache_reset", False)
-            separate_names = data.get("separate_names", False)
+            separate_names = data.get("separate_names", True)
 
             if optimistic_cache_reset:
                 load_parser_model.cache_clear()
@@ -384,6 +385,12 @@ def labeller_bulk_upload():
         return error_response(status=404)
 
 
+def is_valid_dotnum_range(s: str) -> bool:
+    """Checks a str against the format "{digit}..{digit}"""
+
+    return bool(re.fullmatch(r"^\d*\.?\d*(?<!\.)\.\.(?!\.)\d*\.?\d*$", s))
+
+
 @app.route("/labeller/search", methods=["POST"])
 @cross_origin()
 def labeller_search():
@@ -410,12 +417,36 @@ def labeller_search():
                 whole_word = data.get("wholeWord", False)
                 case_sensitive = data.get("caseSensitive", False)
 
+                reserved_char_search = re.search(
+                    RESERVED_LABELLER_SEARCH_CHARS, sentence
+                )
+                reserved_char_match = (
+                    reserved_char_search.group() if reserved_char_search else None
+                )
+
+                # reserve == for id search
+                ids_reserved = []
+                if reserved_char_match in ["=="]:
+                    ids_unique = map(str.strip, list(set(sentence[2:].split(","))))
+                    ids_actual = [
+                        ix
+                        for ix in ids_unique
+                        if ix.isdigit() or is_valid_dotnum_range(ix)
+                    ]
+
+                    for id in ids_actual:
+                        if is_valid_dotnum_range(id):
+                            start, stop = id.split("..")
+                            ids_reserved.extend(range(int(start), int(stop) + 1))
+                        elif id.isdigit():
+                            ids_reserved.append(int(id))
+
                 # preprocess for correct token comparison later
                 sentence_preprocessed = PreProcessor(sentence).sentence
                 # reserve ** or ~~ for wildcard, treat as empty string
                 sentence_cleansed = (
                     " "
-                    if re.search(r"\*\*|~~", sentence_preprocessed)
+                    if reserved_char_match in ["**", "~~"]
                     else sentence_preprocessed
                 )
 
@@ -456,8 +487,10 @@ def labeller_search():
                             if label in labels
                         ]
                     )
-                    if query.search(partial_sentence) or (
-                        partial_sentence == sentence_cleansed
+                    if (
+                        row["id"] in ids_reserved
+                        or query.search(partial_sentence)
+                        or partial_sentence == sentence_cleansed
                     ):
                         indices.append(row["id"])
 
@@ -468,7 +501,7 @@ def labeller_search():
                     cursor.execute(
                         f"""
                         SELECT *
-                        FROM en
+                        FROM {SQL3_DATABASE_TABLE}
                         WHERE id IN ({",".join(["?"] * len(batch))})
                         """,
                         (batch),
@@ -488,7 +521,7 @@ def labeller_search():
                     cursor.execute(
                         f"""
                         SELECT COUNT(*)
-                        FROM en
+                        FROM {SQL3_DATABASE_TABLE}
                         WHERE id IN ({",".join(["?"] * len(batch))})
                         """,
                         (batch),
diff --git a/webtools/src/components/PageTabLabeller/TextInputSubmit/TextInputSubmit.tsx b/webtools/src/components/PageTabLabeller/TextInputSubmit/TextInputSubmit.tsx
@@ -6,16 +6,20 @@ import {
 	type ActionIconProps,
 	Box,
 	Checkbox,
+	Code,
+	Divider,
 	Flex,
 	Group,
 	Kbd,
 	Loader,
 	Menu,
 	MultiSelect,
 	Popover,
+	Stack,
 	Text,
 	TextInput,
 	type TextInputProps,
+	Title,
 	Transition,
 } from "@mantine/core";
 import { getHotkeyHandler, useDisclosure } from "@mantine/hooks";
@@ -26,7 +30,7 @@ import {
 	IconQuestionMark,
 	IconX,
 } from "@tabler/icons-react";
-import { useState } from "react";
+import { type ReactNode, useState } from "react";
 import { useShallow } from "zustand/react/shallow";
 // {{{INTERNAL}}}
 import {
@@ -35,6 +39,68 @@ import {
 	useTabLabellerStore,
 } from "../../../domain";
 
+interface ReservedChars {
+	chars: {
+		value: string;
+		label: string;
+	}[];
+	description: ReactNode;
+}
+
+const reservedCharacters: ReservedChars[] = [
+	{
+		chars: [
+			{ value: "**", label: "(double asterick)" },
+			{ value: "~~", label: "(double tilde)" },
+		],
+		description: (
+			<span>
+				Reserved for wildcard searches. Use this to search the ingredient
+				database without a keyword. For example, input only <Code>**</Code>.
+			</span>
+		),
+	},
+	{
+		chars: [
+			{ value: "== n,n+1", label: "(double equals, comma separated numbers)" },
+		],
+		description: (
+			<span>
+				Reserved for ID matched searches. Use this to directly query the
+				ingredient database row IDs. Supports double dot notation for ID ranges.
+				For example the input <Code>== 1,2,10..14</Code> would be interpreted as{" "}
+				<Code>== 1,2,10,11,12,13,14</Code>.
+			</span>
+		),
+	},
+];
+
+function ReservedCharDescriptor({
+	reservedChars,
+}: {
+	reservedChars: ReservedChars;
+}) {
+	const { chars, description } = reservedChars;
+
+	return (
+		<Stack gap="xs" px="xs" py="sm">
+			<Stack gap="xs">
+				{chars.map(({ label, value }) => (
+					<Group key={`rchar-${label}`} gap={3}>
+						<Kbd>{value}</Kbd>
+						<Text variant="light" size="xs">
+							{label}
+						</Text>
+					</Group>
+				))}
+			</Stack>
+			<Text variant="light" size="sm">
+				{description}
+			</Text>
+		</Stack>
+	);
+}
+
 function ActionIconQuestion(props: ActionIconProps) {
 	const [opened, { close, open }] = useDisclosure(false);
 
@@ -44,7 +110,7 @@ function ActionIconQuestion(props: ActionIconProps) {
 			shadow="md"
 			keepMounted={false}
 			position="bottom-end"
-			width={180}
+			width={350}
 			offset={8}
 		>
 			<Popover.Target>
@@ -59,11 +125,22 @@ function ActionIconQuestion(props: ActionIconProps) {
 				</ActionIcon>
 			</Popover.Target>
 
-			<Popover.Dropdown>
-				<Text variant="light" size="sm">
-					Use double tilde <Kbd>~~</Kbd> or double asterick <Kbd>**</Kbd> to
-					search all records against your filters
-				</Text>
+			<Popover.Dropdown p={0}>
+				<Box px="xs" py="md">
+					<Title order={4} lh={1}>
+						Search Operators
+					</Title>
+				</Box>
+				<Divider />
+				{reservedCharacters.map((reservedChar, i) => (
+					<>
+						<ReservedCharDescriptor
+							key={`rchar-descriptor-${reservedChar.description}`}
+							reservedChars={reservedChar}
+						/>
+						{i + 1 !== reservedCharacters.length && <Divider />}
+					</>
+				))}
 			</Popover.Dropdown>
 		</Popover>
 	);
diff --git a/webtools/src/domain/store/buttonUploadLabellerStore.tsx b/webtools/src/domain/store/buttonUploadLabellerStore.tsx
@@ -259,7 +259,7 @@ export const useUploadNewLabellersStore = create(
 								sentence: "~~",
 								settings: { ...input.settings, sources: [source] },
 							},
-							offsetPage: 1,
+							offsetPage: 0,
 						};
 						getLabellerSearchApi(newInput);
 						updateInput(newInput.input);
diff --git a/webtools/src/domain/store/tabLabellerStore.tsx b/webtools/src/domain/store/tabLabellerStore.tsx
@@ -322,7 +322,7 @@ export const useTabLabellerStore = create(
 		},
 		getLabellerSearchApi: async (opts?: TabLabellerInputProvided) => {
 			const input = opts?.input || get().input;
-			const offset = opts?.offsetPage || 1;
+			const offset = opts?.offsetPage || 0;
 
 			if (input && input.sentence.length !== 0) {
 				set({ loading: true });
