wolfhece 2.2.46__py3-none-any.whl → 2.2.48__py3-none-any.whl

wolfhece/PyCrosssections.py

@@ -2395,7 +2395,7 @@ class crosssections(Element_To_Draw):
         ['right_down']: wolfvertex
         ['cs']      : profile (surcharge de vector)
 
-    Pour le moment, il est possible de lire les fichiers et d'effectuer cerrains traitements (tri selon vecteur, export gltf...).
+    Pour le moment, il est possible de lire les fichiers et d'effectuer certains traitements (tri selon vecteur, export gltf...).
 
     Une instance de cet objet peut être ajouté à une instance graphique WOLF pour affichage.
     Pour ce faire:
@@ -2408,7 +2408,7 @@ class crosssections(Element_To_Draw):
 
     """
 
-    myprofiles:dict['cs':profile, 'index':int, 'left':wolfvertex, 'bed':wolfvertex, 'right':wolfvertex, 'left_down':wolfvertex, 'right_down':wolfvertex]
+    myprofiles:dict[str | int: dict['cs':profile, 'index':int, 'left':wolfvertex, 'bed':wolfvertex, 'right':wolfvertex, 'left_down':wolfvertex, 'right_down':wolfvertex]]
     mygenprofiles:dict
 
     def __init__(self,
@@ -3395,6 +3395,10 @@ class crosssections(Element_To_Draw):
 
         curprof = self[0]
 
+        if curprof.up is None:
+            logging.warning(_('No upstream profile defined for profile %s.')%curprof.myname)
+            return self.myprofiles[curprof.myname]
+
         while curprof.up is not curprof:
             curprof = curprof.up
 
@@ -3405,6 +3409,10 @@ class crosssections(Element_To_Draw):
 
         curprof = self[0]
 
+        if curprof.down is None:
+            logging.warning(_('No downstream profile defined for profile %s.')%curprof.myname)
+            return self.myprofiles[curprof.myname]
+
         while curprof.down is not curprof:
             curprof = curprof.down
 
@@ -3649,6 +3657,10 @@ class crosssections(Element_To_Draw):
         if to_destroy:
             destroy_prepared(vecsupport)
 
+        if len(mysorted)==0:
+            logging.warning(_('No cross-section intersects the support vector!'))
+            return 0
+
         #on trie le résultat en place
         mysorted.sort(key=lambda x:x.s)
 

wolfhece/PyDraw.py

@@ -102,8 +102,10 @@ except ImportError as e:
 
 try:
     from .dike import DikeWolf, InjectorWolf as InjectorDike
+    WOLFPYDIKE_AVAILABLE = True
 except:
     logging.warning(_("Missing package. Install wolfpydike module via pip."))
+    WOLFPYDIKE_AVAILABLE = False
 
 try:
     from .hydrometry.kiwis_wolfgui import hydrometry_wolfgui
@@ -2142,9 +2144,11 @@ class WolfMapViewer(wx.Frame):
     myviewers3d:list[Wolf_Viewer3D]
     mylazdata:list[Wolf_LAZ_Data]
     mydrownings: list[Drowning_victim_Viewer]
-    mydikes: list[DikeWolf]
     mypicturecollections: list[PictureCollection]
-    myinjectors: list[InjectorDike]
+
+    if WOLFPYDIKE_AVAILABLE:
+        mydikes: list[DikeWolf]
+        myinjectors: list[InjectorDike]
 
     mymplfigs:list[MplFigViewer]
 
@@ -2180,8 +2184,11 @@ class WolfMapViewer(wx.Frame):
     active_weir : Weir
     active_laz : Wolf_LAZ_Data
     active_drowning: Drowning_victim_Viewer
-    active_dike : DikeWolf
-    active_injector : InjectorDike
+
+    if WOLFPYDIKE_AVAILABLE:
+        active_dike : DikeWolf
+        active_injector : InjectorDike
+
     active_picturecollection: PictureCollection
     active_alaro: IRM_Alaro
 
@@ -2406,7 +2413,9 @@ class WolfMapViewer(wx.Frame):
         create_acceptability = self.menucreateobj.Append(wx.ID_ANY, _('Create acceptability manager...'), _('Create acceptability manager'))
         create_inbe = self.menucreateobj.Append(wx.ID_ANY, _('Create INBE manager...'), _('Create INBE manager'))
         createdrowning = self.menucreateobj.Append(wx.ID_ANY, _('Create a drowning...'),_('Create a drowning'))
-        createdike = self.menucreateobj.Append(wx.ID_ANY, _('Create dike...'), _('New dike'))
+
+        if WOLFPYDIKE_AVAILABLE:
+            createdike = self.menucreateobj.Append(wx.ID_ANY, _('Create dike...'), _('New dike'))
 
         self.filemenu.AppendSeparator()
 
@@ -2434,7 +2443,9 @@ class WolfMapViewer(wx.Frame):
         addweirs = self.menuaddobj.Append(wx.ID_ANY, _('Add weirs...'), _('Add bridges from directory'))
         addview = self.menuaddobj.Append(wx.ID_ANY, _('Add view...'), _('Add view from project file'))
         adddrowning = self.menuaddobj.Append(wx.ID_ANY, _('Add a drowning result...'),_('Add a drowning result'))
-        adddike = self.menuaddobj.Append(wx.ID_ANY, _('Add dike...'), _('Add dike'))
+
+        if WOLFPYDIKE_AVAILABLE:
+            adddike = self.menuaddobj.Append(wx.ID_ANY, _('Add dike...'), _('Add dike'))
 
         self.precomputed_menu = None
         if self.default_dem != "":
@@ -4295,6 +4306,9 @@ class WolfMapViewer(wx.Frame):
 
     def Onmenudike(self, event: wx.MenuEvent):
 
+        if not WOLFPYDIKE_AVAILABLE:
+            raise ImportError('WolfPyDike not installed -- Please install WolfPyDike from pip to use this feature')
+
         id = event.GetId()
         item = self.menubar.FindItemById(event.GetId())
 
@@ -6182,9 +6196,11 @@ class WolfMapViewer(wx.Frame):
         self.myviewerslaz = []
         self.mylazdata = []
         self.mydrownings = []
+        self.mypicturecollections = []
+
         self.mydikes = []
         self.myinjectors = []
-        self.mypicturecollections = []
+
 
         self.mymplfigs = []
 
@@ -12114,6 +12130,10 @@ class WolfMapViewer(wx.Frame):
 
         elif which.lower() == 'dike':
 
+            if not WOLFPYDIKE_AVAILABLE:
+                logging.error('WolfPyDike module not available - cannot add dike')
+                return -1
+
             curdict = self.mydikes
             curtree = self.myitemsdike
 
@@ -14468,22 +14488,23 @@ class WolfMapViewer(wx.Frame):
             logging.info(_('Activating Drowning victim event : ' + nameitem))
             self.active_drowning = myobj
 
-        elif type(myobj) == DikeWolf:
-            logging.info(_('Activating DikeWolf : ' + nameitem))
-            self.active_dike = myobj
-            if myobj.injector is not None:
-                self.active_injector = myobj.injector
-                logging.info(_('Activating InjectorDike : ' + nameitem))
+        elif WOLFPYDIKE_AVAILABLE:
+            if type(myobj) == DikeWolf:
+                logging.info(_('Activating DikeWolf : ' + nameitem))
+                self.active_dike = myobj
+                if myobj.injector is not None:
+                    self.active_injector = myobj.injector
+                    logging.info(_('Activating InjectorDike : ' + nameitem))
 
-            if ctrl:
-                myobj.show_properties()
+                if ctrl:
+                    myobj.show_properties()
 
-        elif type(myobj) == InjectorDike:
-            logging.info(_('Activating InjectorDike : ' + nameitem))
-            self.active_injector = myobj
+            elif type(myobj) == InjectorDike:
+                logging.info(_('Activating InjectorDike : ' + nameitem))
+                self.active_injector = myobj
 
-            if ctrl:
-                myobj.show_properties()
+                if ctrl:
+                    myobj.show_properties()
 
     def _update_mytooltip(self):
         """ Update the tooltip with the values of the active arrays and results at position x,y """

wolfhece/apps/version.py

@@ -5,7 +5,7 @@ class WolfVersion():
 
         self.major = 2
         self.minor = 2
-        self.patch = 46
+        self.patch = 48
 
     def __str__(self):
 

wolfhece/report/compare_cs_dem.py

@@ -618,7 +618,8 @@ class CSvsDEM():
         if output_file.exists():
             logging.warning(f"Output file {output_file} already exists. It will be overwritten.")
 
-        page = CSvsDEM_IndividualLayout(_("Group {} - Index {} - Cross-Section {}").format(self.index_group, self.index_cs, self._cs.myname))
+        diff_left, diff_right = self.differences
+        page = CSvsDEM_IndividualLayout(_("Group {} - Index {} - CS {} - Delta_left {} - Delta_right {}").format(self.index_group, self.index_cs, self._cs.myname, diff_left, diff_right))
         page.create_report()
         self._complete_report(page)
         page.save_report(output_file)

wolfhece/wolf_array.py

@@ -7698,18 +7698,32 @@ class WolfArray(Element_To_Draw, header_wolf):
 
     def interpolate_on_polygon(self, working_vector: vector,
                                method:Literal["nearest", "linear", "cubic"]="linear",
-                               keep:Literal['all', 'below', 'above'] = 'all'):
+                               keep:Literal['all', 'below', 'above'] = 'all',
+                               rescale:bool = False):
         """
-        Interpolation sous un polygone
+        Interpolation sous un polygone.
 
         L'interpolation a lieu :
           - uniquement dans les mailles sélectionnées si elles existent
           - dans les mailles contenues dans le polygone sinon
 
         On utilise ensuite "griddata" de Scipy pour interpoler les altitudes des mailles
-        depuis les vertices 3D du polygone
+        depuis les vertices 3D du polygone.
+
+        :ATTENTION: l'interpolation est réalisée en réels flottants 64 bits. Le résultat sera ensuite transformé dans le type de la matrice.
+
+        :param working_vector: vector - polygon with z values
+        :param method: interpolation method - 'nearest', 'linear' or 'cubic'
+        :param keep: 'all' to keep all interpolated values
+                     'below' to keep only values below the current value in the array
+                     'above' to keep only values above the current value in the array
+        :param rescale: rescale the input data to [0, 1] for better numerical stability (only for 'linear' and 'cubic' methods)
+        :return: None
         """
 
+        if self.wolftype in [WOLF_ARRAY_FULL_INTEGER, WOLF_ARRAY_FULL_INTEGER8, WOLF_ARRAY_FULL_INTEGER16]:
+            logging.info(_('Your array contains integer data. The interpolation will be done in float and then converted to integer using "np.ceil" operator.'))
+
         if vector.area == 0.:
             logging.error(_('The polygon has no area'))
             return
@@ -7756,7 +7770,7 @@ class WolfArray(Element_To_Draw, header_wolf):
 
         xyz = working_vector.asnparray3d()
 
-        newvalues = griddata(xyz[:, :2], xyz[:, 2], destxy, method=method, fill_value=-99999.)
+        newvalues = griddata(xyz[:, :2], xyz[:, 2], destxy, method=method, fill_value=-99999., rescale = rescale)
 
         if keep == 'all':
             locmask = np.where(newvalues != -99999.)
@@ -7765,6 +7779,11 @@ class WolfArray(Element_To_Draw, header_wolf):
         elif keep == 'above':
             locmask = np.where((newvalues != -99999.) & (newvalues > self.array.data[destij[:, 0], destij[:, 1]]))
 
+        if self.wolftype in [WOLF_ARRAY_FULL_INTEGER, WOLF_ARRAY_FULL_INTEGER8, WOLF_ARRAY_FULL_INTEGER16]:
+            logging.debug(_('Converting interpolated values to integer using "np.ceil" operator'))
+            newvalues = np.ceil(newvalues)
+            newvalues = newvalues.astype(self.array.dtype)
+
         self.array.data[destij[locmask][:, 0], destij[locmask][:, 1]] = newvalues[locmask]
 
     def rasterize_vector_valuebyid(self, working_vector: vector, id,

{wolfhece-2.2.46.dist-info → wolfhece-2.2.48.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: wolfhece
-Version: 2.2.46
+Version: 2.2.48
 Author-email: Pierre Archambeau <pierre.archambeau@uliege.be>
 Project-URL: Homepage, https://uee.uliege.be/hece
 Project-URL: Issues, https://uee.uliege.be/hece
@@ -73,7 +73,7 @@ Requires-Dist: tabulate
 Requires-Dist: ipympl
 Requires-Dist: contextily
 Requires-Dist: pefile
-Requires-Dist: wolfpydike==0.1.3
+Requires-Dist: wolfpydike==0.1.4
 Requires-Dist: PyMuPDF
 Requires-Dist: eccodes
 Requires-Dist: dataframe_image

{wolfhece-2.2.46.dist-info → wolfhece-2.2.48.dist-info}/RECORD

@@ -1,4 +1,3 @@
-wolfhece/ChatwWOLF.py,sha256=B7MkwZiLYjR3OUNBcTIxDnYZzOBFDe52k880KTLUotc,10135
 wolfhece/Coordinates_operations.py,sha256=DSkzJ1Rm4y89I9tuyyAA9mp-EHp9vl5w2qGpNJ-e9qs,8215
 wolfhece/CpGrid.py,sha256=_piG1u-ua7NzWh_PHJYTmxuPJ43ZfeYKNEQgZIJwDJ8,10660
 wolfhece/GraphNotebook.py,sha256=_VZfakR5eXBZE-4Ztv2n12ZDO8zESoeDfCz_9k__T20,31509
@@ -9,8 +8,8 @@ wolfhece/Model1D.py,sha256=-2ibQLscVUsXlcnJWixCIScrBPqJ9BTirmwtGXEKI-4,571155
 wolfhece/MulticriteriAnalysis.py,sha256=vGmkzYagZohNe0XjwGJ6VUXcDPjOt80lNFthXpzxCF0,59572
 wolfhece/PandasGrid.py,sha256=etfVhIHzja4Z1EUY6BcDOKX-w7V-Xou1yaf0NMqmclo,4599
 wolfhece/PyConfig.py,sha256=13DDWjJdohYHwn1uRVHB0s8Jcwq_b9pwcwbAr8NlZyc,19667
-wolfhece/PyCrosssections.py,sha256=iNbDEvC3Ym9jtDzIaxifiNQgYG7vzbnSxUG8P_bcpbk,184104
-wolfhece/PyDraw.py,sha256=Wby6vq-xxQUIF-PoNNBJ4GNgrP6Ug5RX5A3eylYEDP8,744285
+wolfhece/PyCrosssections.py,sha256=kCZBlX3idRY-QqJRHE6MPm4Qxz0AOVublV-JSJYUab8,184622
+wolfhece/PyDraw.py,sha256=bAZGNQrwTJvdPCTnuPZbhj7IvXf9fcPIYofL5U72J9w,744916
 wolfhece/PyGui.py,sha256=z8m4M4Q7DVnRt_bEpEDjUl5x0FMkmsEnH2XHnGKWo14,185336
 wolfhece/PyGuiHydrology.py,sha256=dmBlRO8AljsvCPH6eVt0l9ZLx7g5j7Ubl9Srk7ECwyA,34693
 wolfhece/PyHydrographs.py,sha256=1P5XAURNqCvtSsMQXhOn1ihjTpr725sRsZdlCEhhk6M,3730
@@ -69,7 +68,7 @@ wolfhece/textpillow.py,sha256=7hgfsLYAaE_rNKD-g8xsON8sdWvoV8vbqnGGxIayShE,14137
 wolfhece/tools2d_dll.py,sha256=TfvvmyZUqEZIH0uHwUCJf0bdmCks_AiidDt23Unsp5w,13550
 wolfhece/tools_mpl.py,sha256=gQ3Jg1iuZiecmMqa5Eli2ZLSkttu68VXL8YmMDBaEYU,564
 wolfhece/toolshydrology_dll.py,sha256=cIGyhxV8H5f7GXhDqAamM7uC0W0hQTou3eTkqZdnqBE,5656
-wolfhece/wolf_array.py,sha256=AlVT-1o8D3UWIQcvBUP_oklHOx-vY__9cqdnF8lUtag,588219
+wolfhece/wolf_array.py,sha256=eQGVH_yuJDqLixaho4JZzdPLf6pRf2jgrok86Uy6YUU,589560
 wolfhece/wolf_hist.py,sha256=fTEb60Q4TEwobdZsRU4CFXAId1eOKdWAqF8lnF1xEWc,3590
 wolfhece/wolf_texture.py,sha256=Pt1j_lX74p70Fj3y3qYxYMuN8gghVd8_ih1vFhTIdkA,23884
 wolfhece/wolf_tiles.py,sha256=v-HohqaWuMYdn75XLnA22dlloAG90iwnIqrgnB0ASQ4,10488
@@ -97,8 +96,10 @@ wolfhece/apps/curvedigitizer.py,sha256=lEJJwgAfulrrWQc-U6ij6sj59hWN3SZl4Yu1kQxVz
 wolfhece/apps/hydrometry.py,sha256=lhhJsFeb4zGL4bNQTs0co85OQ_6ssL1Oy0OUJCzhfYE,656
 wolfhece/apps/isocurrent.py,sha256=dagmGR8ja9QQ1gwz_8fU-N052hIw-W0mWGVkzLu6C7I,4247
 wolfhece/apps/splashscreen.py,sha256=EdGDN9NhudIiP7c3gVqj7dp4MWFB8ySizM_tpMnsgpE,3091
-wolfhece/apps/version.py,sha256=4c9wOXhaRWFeLjTaF-KQG_NPeaUiy87evcEae5WZrUE,388
+wolfhece/apps/version.py,sha256=kWKMCvHZwgWn23QbobBBBQzDHb6aopNtwTwORkNiRX8,388
+wolfhece/apps/wolf.ico,sha256=ej-kLodlUrQLsFDxpDhu2Mak7BTuWw6c9NCJaCBOkHI,238323
 wolfhece/apps/wolf.py,sha256=mRnjYsUu4KIsRuamdQWAINFMuwN4eJgMo9erG-hkZ70,729
+wolfhece/apps/wolf2.png,sha256=YXd-MbtlxWRq38l5L6ky4Lg80DFuoZEKdbDwmeSsWqo,24714
 wolfhece/apps/wolf2D.py,sha256=4z_OPQ3IgaLtjexjMKX9ppvqEYyjFLt1hcfFABy3-jU,703
 wolfhece/apps/wolf_logo.bmp,sha256=ruJ4MA51CpGO_AYUp_dB4SWKHelvhOvd7Q8NrVOjDJk,3126
 wolfhece/apps/wolf_logo2.bmp,sha256=pCJFVDn_-rHru6fumazVNM4BqAaobM0Xg0zI0DFWMSQ,5830
@@ -286,7 +287,7 @@ wolfhece/rem/__init__.py,sha256=S2-J5uEGK_VaMFjRUYFIdSScJjZyuXH4RmMmnG3OG7I,19
 wolfhece/report/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 wolfhece/report/common.py,sha256=7nRf1DEXancWsh3V-mqSIqqleaquxn7HemTPTR-8cZ0,21311
 wolfhece/report/compare_arrays.py,sha256=vduobNVgrhWC1LNgCNP7MJ5Dv1Bs-6omUqacbV9-H4w,44050
-wolfhece/report/compare_cs_dem.py,sha256=KujOAaHoHEFyDYrRpAzkUqHgeTFhBpe4xi2efTCz_ig,59517
+wolfhece/report/compare_cs_dem.py,sha256=kojFuXYEUGBB1dgt2W7CROU8ic9idHgQjiHkJHBFPRw,59612
 wolfhece/report/pdf.py,sha256=WNz-uPDASvAtnP-rJaRRdz0nwiSs8XXWe4622FmeVVM,1953
 wolfhece/report/reporting.py,sha256=JUEXovx_S4jpYkJEBU0AC-1Qw2OkkWyV3VAp6iOfSHc,19494
 wolfhece/report/simplesimgpu.py,sha256=nQK1lY9-uSNTA34N9D5U_ftgCVaSMjtq1zPzz2nO7ts,58622
@@ -323,8 +324,8 @@ wolfhece/ui/wolf_multiselection_collapsiblepane.py,sha256=u4C7CXe_bUyGKx7c_Bi0x9
 wolfhece/ui/wolf_times_selection_comparison_models.py,sha256=ORy7fz4dcp691qKzaOZHrRLZ0uXNhL-LIHxmpDGL6BI,5007
 wolfhece/wintab/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 wolfhece/wintab/wintab.py,sha256=8A-JNONV6ujgsgG3lM5Uw-pVgglPATwKs86oBzzljoc,7179
-wolfhece-2.2.46.dist-info/METADATA,sha256=vnmf6HdA8sIxenewSqj2Q-i0azUU9KmlfzywqlRCkVA,2792
-wolfhece-2.2.46.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-wolfhece-2.2.46.dist-info/entry_points.txt,sha256=Jr187pyvA3EeJiQLjZK9yo6mJX7IAn6ygZU9T8qF_gQ,658
-wolfhece-2.2.46.dist-info/top_level.txt,sha256=EfqZXMVCn7eILUzx9xsEu2oBbSo9liWPFWjIHik0iCI,9
-wolfhece-2.2.46.dist-info/RECORD,,
+wolfhece-2.2.48.dist-info/METADATA,sha256=4QTa1R2s7lMNa7-ld4itSLW-MEE3w4erdAGxarCPOsY,2792
+wolfhece-2.2.48.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+wolfhece-2.2.48.dist-info/entry_points.txt,sha256=Jr187pyvA3EeJiQLjZK9yo6mJX7IAn6ygZU9T8qF_gQ,658
+wolfhece-2.2.48.dist-info/top_level.txt,sha256=EfqZXMVCn7eILUzx9xsEu2oBbSo9liWPFWjIHik0iCI,9
+wolfhece-2.2.48.dist-info/RECORD,,

wolfhece/ChatwWOLF.py

@@ -1,200 +0,0 @@
-# Préparation des données pour le modèle ChatWOLF, machine conversationnelle spécialisée dans les questions relatives à WOLF.
-# Les données sont principalemen extraites des fichiers rst de l'aide en ligne mais également des fichiers py de l'API.
-import torch
-print(torch.cuda.is_available())
-
-# Importation des modules nécessaires
-import os
-import re
-import json
-from pathlib import Path
-from typing import List, Dict, Any
-from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain.docstore.document import Document
-from langchain_huggingface import HuggingFaceEmbeddings
-from langchain_community.vectorstores import FAISS
-from langchain.chains.retrieval_qa.base import RetrievalQA
-from langchain_community.llms.huggingface_pipeline import HuggingFacePipeline
-from transformers import pipeline, AutoModelForCausalLM, AutoTokenizer
-from sklearn.model_selection import train_test_split
-from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
-import torch
-from datasets import load_dataset, Dataset
-from transformers import Trainer, TrainingArguments
-import logging
-
-logging.basicConfig(level=logging.INFO)
-logger = logging.getLogger(__name__)
-
-rst_directory = Path("D:/ProgrammationGitLab/HECEPython/docs/source")
-py_directory = Path("D:/ProgrammationGitLab/HECEPython/wolfhece")
-output_directory = Path("D:/ProgrammationGitLab/HECEPython/wolfhece/models/chatwolf")
-output_directory.mkdir(parents=True, exist_ok=True)
-
-# Fonction pour extraire le texte des fichiers rst
-def extract_text_from_rst(file_path: Path) -> str:
-    with open(file_path, 'r', encoding='utf-8') as file:
-        text = file.read()
-    # Nettoyage du texte
-    text = re.sub(r'\.\. _.*?:', '', text)  # Remove references
-    text = re.sub(r'\.\. note::.*?\n\n', '', text, flags=re.DOTALL)  # Remove notes
-    text = re.sub(r'\.\. warning::.*?\n\n', '', text, flags=re.DOTALL)  # Remove warnings
-    text = re.sub(r'\.\. code-block::.*?\n\n', '', text, flags=re.DOTALL)  # Remove code blocks
-    text = re.sub(r'\.\. image::.*?\n\n', '', text, flags=re.DOTALL)  # Remove images
-    text = re.sub(r'\.\. figure::.*?\n\n', '', text, flags=re.DOTALL)  # Remove figures
-    text = re.sub(r'\.\. table::.*?\n\n', '', text, flags=re.DOTALL)  # Remove tables
-    text = re.sub(r'\.\. rubric::.*?\n\n', '', text, flags=re.DOTALL)  # Remove rubrics
-    text = re.sub(r'\.\. sidebar::.*?\n\n', '', text, flags=re.DOTALL)  # Remove sidebars
-    text = re.sub(r'\.\. literalinclude::.*?\n\n', '', text, flags=re.DOTALL)  # Remove literal includes
-    text = re.sub(r'\.\. math::.*?\n\n', '', text, flags=re.DOTALL)  # Remove math
-    text = re.sub(r'\.\. raw::.*?\n\n', '', text, flags=re.DOTALL)  # Remove raw
-    text = re.sub(r'\.\. toctree::.*?\n\n', '', text, flags=re.DOTALL)  # Remove toctree
-    text = re.sub(r'\.\. index::.*?\n\n', '', text, flags=re.DOTALL)  # Remove index
-    text = re.sub(r'\.\. glossary::.*?\n\n', '', text, flags=re.DOTALL)  # Remove glossary
-    text = re.sub(r'\.\. footnote::.*?\n\n', '', text, flags=re.DOTALL)  # Remove footnotes
-    text = re.sub(r'\.\. citation::.*?\n\n', '', text, flags=re.DOTALL)  # Remove citations
-    text = re.sub(r'\.\. epigraph::.*?\n\n', '', text, flags=re.DOTALL)  # Remove epigraphs
-    text = re.sub(r'\.\. highlight::.*?\n\n', '', text, flags=re.DOTALL)  # Remove highlights
-    text = re.sub(r'\.\. hlist::.*?\n\n', '', text, flags=re.DOTALL)  # Remove hlists
-    text = re.sub(r'\.\. csv-table::.*?\n\n', '', text, flags=re.DOTALL)  # Remove csv-tables
-    text = re.sub(r'\.\. list-table::.*?\n\n', '', text, flags=re.DOTALL)  # Remove list-tables
-    text = re.sub(r'\.\. contents::.*?\n\n', '', text, flags=re.DOTALL)  # Remove contents
-    text = re.sub(r'\.\. include::.*?\n\n', '', text, flags=re.DOTALL)  # Remove includes
-    text = re.sub(r'\.\. admonition::.*?\n\n', '', text, flags=re.DOTALL)  # Remove admonitions
-    text = re.sub(r'\.\. note::.*?\n\n', '', text, flags=re.DOTALL)  # Remove notes
-    text = re.sub(r'\.\. tip::.*?\n\n', '', text, flags=re.DOTALL)  # Remove tips
-    text = re.sub(r'\.\. important::.*?\n\n', '', text, flags=re.DOTALL)  # Remove importants
-    text = re.sub(r'\.\. caution::.*?\n\n', '', text, flags=re.DOTALL)  # Remove cautions
-    text = re.sub(r'\.\. seealso::.*?\n\n', '', text, flags=re.DOTALL)  # Remove seealso
-
-    return text
-
-def scan_files() -> List[Path]:
-    # Scan all files and extract text
-    documents = []
-    for rst_file in rst_directory.rglob("*.rst"):
-        text = extract_text_from_rst(rst_file)
-        if text.strip():  # Only add non-empty documents
-            documents.append(Document(page_content=text, metadata={"source": str(rst_file)}))
-            logger.info(f"Extracted text from {rst_file}")
-    for py_file in py_directory.rglob("*.py"):
-        with open(py_file, 'r', encoding='utf-8') as file:
-            text = file.read()
-            if text.strip():  # Only add non-empty documents
-                documents.append(Document(page_content=text, metadata={"source": str(py_file)}))
-                logger.info(f"Extracted text from {py_file}")
-    logger.info(f"Total documents extracted: {len(documents)}")
-    return documents
-
-def split_and_prepare_data(documents: List[Document]) -> None:
-    # Split documents into smaller chunks
-    text_splitter = RecursiveCharacterTextSplitter(
-        chunk_size=1000,
-        chunk_overlap=100,
-        length_function=len
-    )
-    texts = text_splitter.split_documents(documents)
-    logger.info(f"Total text chunks created: {len(texts)}")
-    # Save texts to JSONL for dataset creation
-    jsonl_path = output_directory / "chatwolf_data.jsonl"
-    with open(jsonl_path, 'w', encoding='utf-8') as f:
-        for text in texts:
-            json.dump({"text": text.page_content}, f)
-            f.write('\n')
-    logger.info(f"Saved text chunks to {jsonl_path}")
-    return texts, jsonl_path
-
-def train_model():
-    # Load dataset
-    dataset = load_dataset('json', data_files=str(jsonl_path))['train']
-    # Split dataset into training and validation sets
-    train_test_split = dataset.train_test_split(test_size=0.1)
-    train_dataset = train_test_split['train']
-    eval_dataset = train_test_split['test']
-    logger.info(f"Training dataset size: {len(train_dataset)}")
-    logger.info(f"Validation dataset size: {len(eval_dataset)}")
-    # Define model and tokenizer
-    model_name = "gpt2"
-    tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=True)
-    if tokenizer.pad_token is None:
-        tokenizer.pad_token = tokenizer.eos_token
-    model = AutoModelForCausalLM.from_pretrained(model_name)
-    # Define training arguments
-    training_args = TrainingArguments(
-        output_dir=output_directory / "output",
-        eval_strategy="epoch",
-        num_train_epochs=3,
-        per_device_train_batch_size=1,
-        per_device_eval_batch_size=1,
-        save_strategy="epoch",
-        logging_dir=output_directory / "logs",
-        logging_steps=10,
-        save_total_limit=2,
-        fp16=False,  # Set to False to avoid FP16 errors on unsupported hardware
-        load_best_model_at_end=True,
-    )
-    # Define data collator
-    def tokenize_function(examples):
-        return tokenizer(examples["text"], padding="max_length", truncation=True, max_length=512)
-    train_dataset = train_dataset.map(tokenize_function, batched=True)
-    eval_dataset = eval_dataset.map(tokenize_function, batched=True)
-    train_dataset.set_format(type='torch', columns=['input_ids', 'attention_mask'])
-    eval_dataset.set_format(type='torch', columns=['input_ids', 'attention_mask'])
-    # Define data collator for causal language modeling
-    from transformers import DataCollatorForLanguageModeling
-    data_collator = DataCollatorForLanguageModeling(
-        tokenizer=tokenizer,
-        mlm=False,
-    )
-    # Initialize Trainer
-    trainer = Trainer(
-        model=model,
-        args=training_args,
-        train_dataset=train_dataset,
-        eval_dataset=eval_dataset,
-        tokenizer=tokenizer,
-        data_collator=data_collator,
-    )
-    # Train the model
-    trainer.train()
-    # Save the fine-tuned model
-    trainer.save_model(output_directory / "chatwolf_model")
-    logger.info(f"Saved fine-tuned model to {output_directory / 'chatwolf_model'}")
-    return model, tokenizer
-
-def load_model_and_tokenizer():
-    model = AutoModelForCausalLM.from_pretrained(output_directory / "chatwolf_model")
-    tokenizer = AutoTokenizer.from_pretrained("gpt2", use_fast=True)
-    if tokenizer.pad_token is None:
-        tokenizer.pad_token = tokenizer.eos_token
-    return model, tokenizer
-
-documents = scan_files()
-texts, jsonl_path = split_and_prepare_data(documents)
-
-if False:
-    model, tokenizer = train_model()
-else:
-    model, tokenizer = load_model_and_tokenizer()
-
-
-# Create embeddings and vector store
-embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
-vector_store = FAISS.from_documents(texts, embeddings)
-vector_store.save_local(str(output_directory / "faiss_index"))
-logger.info(f"Saved FAISS index to {output_directory / 'faiss_index'}")
-# Create retrieval QA chain
-llm_pipeline = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, temperature=0.7, top_p=0.9, repetition_penalty=1.2)
-hf_llm = HuggingFacePipeline(pipeline=llm_pipeline)
-qa_chain = RetrievalQA.from_chain_type(llm=hf_llm, chain_type="stuff", retriever=vector_store.as_retriever())
-# Save the QA chain
-import pickle
-with open(output_directory / "qa_chain.pkl", 'wb') as f:
-    pickle.dump(qa_chain, f)
-logger.info(f"Saved QA chain to {output_directory / 'qa_chain.pkl'}")
-# Example usage of the QA chain
-def answer_question(question: str) -> str:
-    return qa_chain.run(question)
-example_question = "How to create a new map in WOLF?"
-answer = answer_question(example_question)
-logger.info(f"Question: {example_question}\nAnswer: {answer}")