s495731
/
KWT-2024
forked from bfijalkowski/KWT-2024
1
0
Fork 0
Code Pull Requests Activity

first task

This commit is contained in:
Adam Stelmaszyk 2024-04-13 13:33:54 +02:00
parent 71ca3b66ed
commit 22d22f8f7f
2 changed files with 159 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

182
lab/lab_01.ipynb
View File

@ -52,7 +52,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 1, "execution_count": 2,
"id": "narrow-romantic", "id": "narrow-romantic",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -71,7 +71,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 2, "execution_count": 3,
"id": "indonesian-electron", "id": "indonesian-electron",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -82,7 +82,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": 4,
"id": "compact-trinidad", "id": "compact-trinidad",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -92,7 +92,7 @@
"['Press the ENTER button']" "['Press the ENTER button']"
] ]
}, },
"execution_count": 3, "execution_count": 4,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -119,7 +119,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 4, "execution_count": 5,
"id": "exposed-daniel", "id": "exposed-daniel",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -139,7 +139,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 5, "execution_count": 6,
"id": "serial-velvet", "id": "serial-velvet",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -149,7 +149,7 @@
"['Press the ENTER button', 'Press the ENTER key']" "['Press the ENTER button', 'Press the ENTER key']"
] ]
}, },
"execution_count": 5, "execution_count": 6,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -176,7 +176,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 6, "execution_count": 9,
"id": "every-gibson", "id": "every-gibson",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -186,7 +186,7 @@
"[]" "[]"
] ]
}, },
"execution_count": 6, "execution_count": 9,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -213,13 +213,26 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 7, "execution_count": 11,
"id": "protected-rings", "id": "protected-rings",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"data": {
"text/plain": [
"['Press the ENTER button', 'Press the ENTER key']"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def tm_lookup(sentence):\n", "def tm_lookup(sentence):\n",
" return ''" " return [entry[1] for entry in translation_memory if entry[0].lower() == sentence.lower()]\n",
"\n",
"tm_lookup('Wciśnij przycisk ENTER')"
] ]
}, },
{ {
@ -232,17 +245,17 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 18, "execution_count": 8,
"id": "severe-alloy", "id": "severe-alloy",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"data": { "data": {
"text/plain": [ "text/plain": [
"''" "[]"
] ]
}, },
"execution_count": 18, "execution_count": 8,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -261,13 +274,29 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 11, "execution_count": 16,
"id": "structural-diesel", "id": "structural-diesel",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"data": {
"text/plain": [
"['Press the ENTER button', 'Press the ENTER key']"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"import string\n",
"\n",
"def tm_lookup(sentence):\n", "def tm_lookup(sentence):\n",
" return ''" " sentence = sentence.translate(str.maketrans('', '', string.punctuation))\n",
" return [entry[1] for entry in translation_memory if entry[0].lower() == sentence.lower()]\n",
"\n",
"tm_lookup('Wciśnij przycisk [ENTER]')"
] ]
}, },
{ {
@ -280,17 +309,17 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 12, "execution_count": 17,
"id": "brief-senegal", "id": "brief-senegal",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"data": { "data": {
"text/plain": [ "text/plain": [
"''" "[]"
] ]
}, },
"execution_count": 12, "execution_count": 17,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -317,13 +346,66 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 14, "execution_count": 26,
"id": "mathematical-customs", "id": "mathematical-customs",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Suggestion:\n"
]
},
{
"data": {
"text/plain": [
"['System restart required']"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def differenceThanNotBiggerThanOneElement(firstSentence, secondSentence):\n",
" firstSentenceList = firstSentence.lower().split()\n",
" secondSentenceList = secondSentence.lower().split()\n",
"\n",
" diffNumber = 0\n",
"\n",
" for i in range(len(firstSentenceList)):\n",
" if(firstSentenceList[i] != secondSentenceList[i]):\n",
" diffNumber=diffNumber+1\n",
" if(diffNumber > 2):\n",
" return False\n",
"\n",
" return True\n",
"\n",
"def tm_lookup(sentence):\n", "def tm_lookup(sentence):\n",
" return ''" " sentence = sentence.translate(str.maketrans('', '', string.punctuation))\n",
"\n",
" exactMatchList = [entry[1] for entry in translation_memory if entry[0].lower() == sentence.lower()]\n",
"\n",
" if(len(exactMatchList) == 0):\n",
" diffMatchList = [entry[1] for entry in translation_memory if differenceThanNotBiggerThanOneElement(entry[0], sentence)]\n",
"\n",
" if(len(diffMatchList) > 0):\n",
" print('Suggestion:')\n",
" return diffMatchList\n",
"\n",
" else:\n",
" return exactMatchList\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
" \n",
"\n",
"tm_lookup('Wymagane ponowne uruchomienie maszyny')"
] ]
}, },
{ {
@ -344,7 +426,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 15, "execution_count": 27,
"id": "humanitarian-wrong", "id": "humanitarian-wrong",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -362,7 +444,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 16, "execution_count": 28,
"id": "located-perception", "id": "located-perception",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -374,7 +456,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 17, "execution_count": 29,
"id": "advised-casting", "id": "advised-casting",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -384,7 +466,7 @@
"[('przycisk', 'button'), ('drukarka', 'printer')]" "[('przycisk', 'button'), ('drukarka', 'printer')]"
] ]
}, },
"execution_count": 17, "execution_count": 29,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -406,7 +488,7 @@
"id": "defensive-fifteen", "id": "defensive-fifteen",
"metadata": {}, "metadata": {},
"source": [ "source": [
"Odpowiedź:" "Odpowiedź: O(n * m)"
] ]
}, },
{ {
@ -419,13 +501,27 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 19, "execution_count": 34,
"id": "original-tunisia", "id": "original-tunisia",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"data": {
"text/plain": [
"[]"
]
},
"execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def glossary_lookup(sentence):\n", "def glossary_lookup(sentence):\n",
" return ''" " sentence_words = sentence.lower().split()\n",
" return [entry for entry in glossary if entry[0] in sentence_words]\n",
"\n",
"glossary_lookup('Każda Drukarka posiada Przycisk wznowienia drukowania')"
] ]
}, },
{ {
@ -438,13 +534,27 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 20, "execution_count": 35,
"id": "adolescent-semiconductor", "id": "adolescent-semiconductor",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [
{
"data": {
"text/plain": [
"[('przycisk', 'button'), ('drukarka', 'printer')]"
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def glossary_lookup(sentence):\n", "def glossary_lookup(sentence):\n",
" return ''" " sentence_words = set(sentence.lower().split())\n",
" return [entry for entry in glossary if entry[0] in sentence_words]\n",
"\n",
"glossary_lookup('Każda Drukarka posiada Przycisk wznowienia drukowania')"
] ]
} }
], ],
@ -467,7 +577,7 @@
"name": "python", "name": "python",
"nbconvert_exporter": "python", "nbconvert_exporter": "python",
"pygments_lexer": "ipython3", "pygments_lexer": "ipython3",
"version": "3.8.10" "version": "3.11.7"
}, },
"subtitle": "1. Podstawowe techniki wspomagania tłumaczenia", "subtitle": "1. Podstawowe techniki wspomagania tłumaczenia",
"title": "Komputerowe wspomaganie tłumaczenia", "title": "Komputerowe wspomaganie tłumaczenia",

14
lab/lab_02.ipynb
View File

@ -86,7 +86,19 @@
"outputs": [], "outputs": [],
"source": [ "source": [
"def ice_lookup(sentence, prev_sentence, next_sentence):\n", "def ice_lookup(sentence, prev_sentence, next_sentence):\n",
" return []" " sentence = sentence.lower()\n",
" prev_sentence = prev_sentence.lower()\n",
" next_sentence = next_sentence.lower()\n",
" for index in range(len(translation_memory)):\n",
" if index == 0:\n",
" if len(translation_memory) > 1:\n",
" if sentence == translation_memory[index] and next_sentence == translation_memory[index+1]:\n",
" return True\n",
" else if sentence == translation_memory[index]:\n",
" return True\n",
" \n",
" \n",
" # return [entry[1] for entry in translation_memory if entry[0] == sentence]"
] ]
}, },
{ {