finální update 4. 12. 2024

DataRozhlas · Dec 4, 2024 · bc4aa57 · bc4aa57
1 parent 37614da
commit bc4aa57
Show file tree

Hide file tree

Showing 10 changed files with 272,021 additions and 286,078 deletions.
diff --git a/002_cd_cisteni.ipynb b/002_cd_cisteni.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": 1,
    "id": "e27c863b-fd97-4b32-b4b3-0825cb67f49e",
    "metadata": {},
    "outputs": [],
@@ -15,7 +15,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 2,
    "id": "ad771ad2-9650-42bf-a22b-b9249d883ccd",
    "metadata": {},
    "outputs": [],
@@ -30,7 +30,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 3,
    "id": "4d3db07f-3600-48ac-b158-70ea455b365a",
    "metadata": {},
    "outputs": [],
@@ -116,7 +116,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 4,
    "id": "e297e728-2198-4215-af23-6cf12ab2802f",
    "metadata": {
     "scrolled": true
@@ -126,8 +126,8 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "cd_2024-11-26.parquet\n",
-      "2024-11-26: 84\n"
+      "cd_2024-12-01.parquet\n",
+      "2024-12-01: 825\n"
      ]
     }
    ],

diff --git a/004_rj_cisteni.ipynb b/004_rj_cisteni.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 1,
    "id": "dd9891c3-234b-46f1-8aa5-b37b77890ab0",
    "metadata": {},
    "outputs": [],
@@ -15,7 +15,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 2,
    "id": "6bf4b9fb-fdf5-40e9-a492-9cfc761a1ec4",
    "metadata": {},
    "outputs": [],
@@ -79,7 +79,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 3,
    "id": "4001b228-a1f4-40d8-97b8-e8152ce730dc",
    "metadata": {},
    "outputs": [
@@ -93,7 +93,7 @@
       "2024-11-03: 0\n",
       "2024-11-04: 0\n",
       "2024-11-05: 0\n",
-      "2024-11-26: 673\n"
+      "2024-12-01: 199\n"
      ]
     }
    ],

diff --git a/006_le_cisteni.ipynb b/006_le_cisteni.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 1,
    "id": "b0a1dca1-f4fa-4174-aa26-f2b8cbb5ea79",
    "metadata": {},
    "outputs": [],
@@ -15,7 +15,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 2,
    "id": "dcfe8419-4463-470c-aed0-b72de890da5e",
    "metadata": {},
    "outputs": [],
@@ -103,15 +103,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 3,
    "id": "ace7b62e-9cf2-4268-aac8-29a94271508f",
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "2024-11-26: 254\n"
+      "2024-12-01: 291\n"
      ]
     }
    ],

diff --git a/008_ar_cisteni.ipynb b/008_ar_cisteni.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 1,
    "id": "052ba83e-db70-4a7d-bb1d-df25d257950c",
    "metadata": {},
    "outputs": [],
@@ -15,21 +15,10 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 2,
    "id": "b595d7f5-6aab-41dd-8a96-32bbef2c4b90",
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "<>:11: SyntaxWarning: invalid escape sequence '\\d'\n",
-      "<>:11: SyntaxWarning: invalid escape sequence '\\d'\n",
-      "C:\\Users\\micha\\AppData\\Local\\Temp\\ipykernel_15272\\1036985530.py:11: SyntaxWarning: invalid escape sequence '\\d'\n",
-      "  datum = re.search(\"\\d{1,2}\\.\\s\\w{4,15}\\s20\\d{2}\", radek).group(0)\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "def oscrapuj_ar(slozka, soubor):\n",
     "    with open(os.path.join(slozka,soubor), \"r\", encoding=\"utf-8\") as spojeni:\n",
@@ -41,7 +30,7 @@
     "        for index, radek in enumerate(spojeni.split(\">\")):\n",
     "            if len(radek) < 1000:\n",
     "                if re.search(r\"\\d{1,2}\\.\\s\\w{4,15}\\s20\\d{2}\", radek):\n",
-    "                    datum = re.search(\"\\d{1,2}\\.\\s\\w{4,15}\\s20\\d{2}\", radek).group(0)\n",
+    "                    datum = re.search(r\"\\d{1,2}\\.\\s\\w{4,15}\\s20\\d{2}\", radek).group(0)\n",
     "                    datum = datum.replace(\"leden\",\"1.\").replace(\"únor\",\"2.\").replace(\"březen\",\"3.\").replace(\"duben\",\"4.\").replace(\"květen\",\"5.\").replace(\"červen\",\"6.\").replace('červenec','7.').replace('srpen','8.').replace('září','9.').replace('říjen','10.').replace('listopad','11.').replace('prosinec','12.').replace(' ','')\n",
     "                elif ('class=\"departure\"' in radek) or ('</html' in radek):\n",
     "                    if spoj:\n",
@@ -81,7 +70,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 3,
    "id": "6e76aab9-797d-4448-9284-a6db0dea9b71",
    "metadata": {},
    "outputs": [
@@ -96,7 +85,7 @@
       "2024-11-04: 0\n",
       "2024-11-05: 0\n",
       "2024-11-06: 0\n",
-      "2024-11-26: 659\n"
+      "2024-12-01: 686\n"
      ]
     }
    ],