from django.conf import settings from django.conf.urls.static import static from django.contrib import admin from django.contrib.auth.views import PasswordResetView, PasswordResetConfirmView # class-based views from django.urls import include, path, re_path from troggle.core.views import statistics, survex from troggle.core.views.auth import expologin, expologout from troggle.core.views.caves import ( cave3d, cave_debug, caveindex, cavepage, caveQMs, caveslist, caveslugfwd, edit_cave, edit_entrance, entranceindex, entrancetags, expo_kml, expo_kmz, get_entrances, qm, ) from troggle.core.views.cave_kataster import kataster, fix from troggle.core.views.drawings import dwgallfiles, dwgfilesingle from troggle.core.views.editor_helpers import image_selector, new_image_form from troggle.core.views.expo import ( editexpopage, expofiles_redirect, expofilessingle, expopage, indexpage, map, mapfile, mediapage, pubspage, spider, ) from troggle.core.views.logbook_edit import logbookedit from troggle.core.views.logbooks import ( Expeditions_jsonListView, Expeditions_tsvListView, QMs_jsonListView, expedition, get_logbook_entries, get_people, logbookentry, logentrydelete, logreport, notablepersons, people_ids, person, personexpedition, ) from troggle.core.views.other import controlpanel, exportlogbook, frontpage, todos from troggle.core.views.prospect import prospecting from troggle.core.views.user_registration import register, newregister, reset_done, ExpoPasswordResetForm from troggle.core.views.scans import allscans, cavewallets, scansingle, walletslistperson, walletslistyear from troggle.core.views.signup import signup, signupok from troggle.core.views.uploads import dwgupload, expofilerename, gpxupload, photoupload from troggle.core.views.wallets_edit import walletedit # from troggle.core.views.user_registration import SignUpView # Warning: a Class-based View """This sets the actualurlpatterns[] and urlpatterns[] lists which django uses to resolve urls - in both directions as these are declarative. HOW THIS WORKS This is a "url dispatcher" - something needed by every web framework. re_path( , , ) Django also provides the reverse function: given an an object, provide the URL which is important to writing code for the webapp. So the URL dispatch is declarative. But this means that two URLs should NOT go to the same python target function, (or only if the target name is different) The API urls return TSV or JSON and are new in July 2020. CONVENTION Unlike most DJango projects, we have the convention that we do NOT have a terminal slash on the URL of a page which is generated. (Generated pages are neither files not directories.) So the url is "/dwgfiles" not "/dwgfiles/" everywhere, and similarly for "/walletedit" etc. This is important in troggle because we do NOT universally use the reverse() function or the url() function to get the URL from the declarations in this url.py file. The reason is that url() requires quite a wide knowledge of troggle, whereas explicit urls can be done by beginner maintainers and work (but do add to future maintenance). NOTE - The admin and logout paths need to stay using re_path() as they have to be locked to the start. - admin and login forms are provided by Django so we have to fit in to their conventions. - The final _edit and CATCHALL also have to use re_path(). (?P) This strange syntax is not a Django funny, it is standard (but obscure) python. It denotes a "Named Capturing Group" https://docs.python.org/3/howto/regex.html#non-capturing-and-named-groups https://dnmtechs.com/understanding-the-django-regular-expression-p-in-python-3/ https://www.regular-expressions.info/named.html e.g. the regular expression: r'^(?P.*)/(?P[^/]+)_cave_edit/$' Many of these patterns do not work because troggle spent many years broken and we have not yet restored all the functions. Some may have never been fully implemented in the first place and what they were intended to provide is obscure. Some short names such as indxal.htm date from the Archimedes-era when filenames had to be less than 10-chars long. """ todo = ''' - Replace more re_path() with modern and simpler path(). Careful: some have to stay as re_path() - Test VERY CAREFULLY for each change. It is fragile. ''' # WHen running on the server, apache intercepts all the /expofiles/ files so troggle never sees them, # so the "content type" is set by whatever apache thinks it should be. Which means .gpx files # get treated as XML and the web browser fails to do anything useful if settings.EXPOFILESREMOTE: expofilesurls = [ path('', expofiles_redirect, name="expofiles_redirect"), # to http://expo.survex.com/expofiles path('', expofiles_redirect, {'filepath': ""}, name="expofiles_redirect"), #should not have duplicate urls as reverse() then fails ] else: expofilesurls = [ path('', expofilessingle, {'filepath': ""}, name="single"), path('', expofilessingle, name="single"), # local copy of EXPOFILES, but should not have duplicate urls as reverse() then fails ] trogglepatterns = [ # NOT intercepted by apache. Needs to come first to prevent our troggle middleware trying to "fix" it. re_path(r'^site_media/(?P.*)$', mediapage, {'doc_root': settings.MEDIA_ROOT}, name="mediapage"), # MEDIA_ROOT: CSS and JS path('pubs.htm', pubspage, name="pubspage"), # ~165 hrefs to this url in expoweb files #path('', indexpage, name="indexpage"), # ~1,212 hrefs to this url in expoweb files path('index.htm', indexpage, name="indexpage"), # ~1,212 hrefs to this url in expoweb files path('expofiles/', include(expofilesurls)), # intercepted by Apache, if it is running. path('expofiles', include(expofilesurls)), # curious interaction with the include() here, not just a slash problem. re_path(r'^(.*)_edit_edit$', spider, name="spider"), # web spider. Intercept and manage it. re_path(r'^caves$', caveindex, name="caveindex"), re_path(r'^indxal.htm$', caveindex, name="caveindex"), # ~420 hrefs to this url in expoweb files re_path(r'^people/?$', notablepersons, name="notablepersons"), path('people_ids', people_ids, name="people_ids"), path('caveslist', caveslist, name="caveslist"), path('entrances', entranceindex, name="entranceindex"), path('enttags', entrancetags, name="entrancetags"), re_path(r'^admin/doc/', include('django.contrib.admindocs.urls')), # needs docutils Python module (http://docutils.sf.net/). path('admin/', admin.site.urls), # includes admin login & logout urls & /admin/jsi18n/ NOTE TERMINAL SLASH # Uploads - uploading a file path('walletedit', walletedit, name='walletedit'), # not just an upload, also edit metadata path('walletedit/', walletedit, name='walletedit'), # path=2020#01 path('photoupload', photoupload, name='photoupload'), # restricted to current year path('photoupload/', photoupload, name='photoupload'), # restricted to current year path('gpxupload', gpxupload, name='gpxupload'), # restricted to current year path('gpxupload/', gpxupload, name='gpxupload'), # restricted to current year path('dwgupload/', dwgupload, name='dwgupload'), path('dwgupload', dwgupload, name='dwgupload'), path('dwguploadnogit', dwgupload, {'gitdisable': 'yes'}, name='dwguploadnogit'), # used in testing path('dwguploadnogit/', dwgupload, {'gitdisable': 'yes'}, name='dwguploadnogit'), # used in testing path('logbookedit/', logbookedit, name='logbookedit'), path('logbookedit/', logbookedit, name='logbookedit'), path('signup', signup, name='signup'), path('signupok', signupok, name='signupok'), # Renaming an uploaded file path('expofilerename/', expofilerename, name='expofilerename'), # see https://docs.djangoproject.com/en/dev/topics/auth/default/tiny # The URLs provided by include('django.contrib.auth.urls') are: # # accounts/login/ [name='login'] # accounts/logout/ [name='logout'] # accounts/password_change/ [name='password_change'] # accounts/password_change/done/ [name='password_change_done'] # accounts/password_reset/ [name='password_reset'] # accounts/password_reset/done/ [name='password_reset_done'] # accounts/reset/// [name='password_reset_confirm'] # accounts/reset/done/ [name='password_reset_complete'] # BUT many of these are set up by opinionated Django even if 'django.contrib.auth.urls' is NOT included. # Some overlap with 'admin.site.urls' needs to be investigated. # admin.site.urls is urls() which maps to get_urls() which is a function declared # in django/contrib/admin/sites.py which for me is # /home/philip/expo/troggle/.venv/lib/python3.xx/site-packages/django/contrib/admin/sites.py # setting LOGIN_URL = '/accounts/login/' is default. # NB setting url pattern name to 'login' instead of 'expologin' with override Django, see https://docs.djangoproject.com/en/dev/topics/http/urls/#naming-url-patterns path('accounts/logout/', expologout, name='expologout'), # same as in django.contrib.auth.urls path('accounts/login/', expologin, name='expologin'), # same as in django.contrib.auth.urls path("accounts/register/", register, name="re_register"), # overriding django.contrib.auth.urls path("accounts/register/", register, name="register"), # overriding django.contrib.auth.urls path("accounts/newregister/", newregister, name="newregister"), path("accounts/reset/done/", reset_done, name="password_reset_done"), # overriding django.contrib.auth.urls path('accounts/password_reset/', PasswordResetView.as_view(form_class=ExpoPasswordResetForm), name='password_reset'), path('accounts/reset///', PasswordResetConfirmView.as_view(), name="password_reset_confirm"), path('accounts/', include('django.contrib.auth.urls')), # see line 109 in this file NB initial "/accounts/" in URL path('person/', person, name="person"), path('personexpedition//', personexpedition, name="personexpedition"), # Expedition master page & API exports re_path(r'^expedition/(\d+)$', expedition, name="expedition"), re_path(r'^api/expeditions_tsv$', Expeditions_tsvListView.as_view()), re_path(r'^api/expeditions_json$', Expeditions_jsonListView.as_view()), re_path(r'^api/QMs_json$', QMs_jsonListView.as_view()), # Logbook entries re_path(r'^logbookentry/(?P.*)/(?P.*)/?$', logbookentry,name="logbookentry"), re_path(r'^logbook$', exportlogbook, name='exportlogbook'), path('logreport/', logreport, name='logreport'), path('logentrydelete/', logentrydelete, name='logentrydelete'), # Internal. editfile.html template uses these internally re_path(r'^getPeople/(?P.*)', get_people, name = "get_people"), re_path(r'^getLogBookEntries/(?P.*)', get_logbook_entries, name = "get_logbook_entries"), re_path(r'^getEntrances/(?P.*)', get_entrances, name = "get_entrances"), # Cave description pages path('cave/', caveslugfwd, name="caveslugfwd"), path('cave_debug', cave_debug, name="cave_debug"), path('kataster/', kataster, name="kataster"), # for renaming a cave from e.g. 1623-2005-05 to 1623-264 path('kataster', kataster, name="kataster"), # illustrative placeholder for kataster renaming path('fix/', fix, name="fix"), # one-off fix misplaced images and descriptive files re_path(r'^newcave/$', edit_cave, name="newcave"), re_path(r'^cave/3d/(?P[^/]+).3d$', cave3d, name="cave3d"), # Edit caves and entrances re_path(r'^(?P.*)/(?P[^/]+)_cave_edit/$', edit_cave, name="edit_cave"), # edit_cave needed by cave.html template for url matching re_path(r'^(?P[^/]+):(?P[^:]+)_entrance_edit', edit_entrance, name = "editentrance"), #edit existing entrance re_path(r'^(?P.*)/(?P[^/]+)_entrance_new$', edit_entrance, name = "newentrance"), # new entrance for a cave re_path(r'^(.*)_edit$', editexpopage, name="editexpopage"), # only happens if not a cave or entrance _edit page re_path(r'^(?P162\d)(?P.*)$', cavepage, name="cavepage"), # shorthand /1623/264 or 1623/161/top.htm or 1623/161/i/stuff.jpg # Note that urls eg '/1623/161/l/rl89a.htm' are handled by cavepage which redirects them to 'expopage' # Note that _edit$ for a cave description page in a subfolder e.g. /1623/204/204.html_edit gets caught here and breaks with 404 # These re-enable archaic URLs which are in old web pages which may still be public on other servers - old # path('smkridge/', cavepagefwd, {'karea': "1623"}, name="cavepagefwd"), # path('expo/smkridge/', cavepagefwd, {'karea': "1623"}, name="cavepagefwd"), # Archaic, kept. This /expo/ prefix only works for expoweb HTML pages not troggle pages path('expo/', expopage, name="expopage"), # System admin and monitoring path('statistics', statistics.stats, name="stats"), path('stats', statistics.stats, name="stats"), path('pathsreport', statistics.pathsreport, name="pathsreport"), path('dataissues', statistics.dataissues, name="dataissues"), path('therionissues', statistics.therionissues, name="therionissues"), path('surveximport', statistics.surveximport, name="surveximport"), path('survexdebug', statistics.survexdebug, name="survexdebug"), path('stations', statistics.stations, name="stations"), path('aliases/',statistics.aliases, name="aliases"), path('troggle', frontpage, name="frontpage"), # control panel. Shows recent actions. path('todo/', todos, name="todos"), path('controlpanel', controlpanel, name="controlpanel"), # The survexfile pages path('survexdir', survex.survexdir, name="survexdir"), path('survexfile', survex.survexcavesingle, {'cave_shortname': ''}, name="survexcavessingle"), path('survexfile/caves', survex.survexcaveslist, name="survexcaveslist"), path('survexfile/.svx', survex.svx, name="svx"), path('survexfile/.3d', survex.threed, name="threed"), path('survexfile/.log', survex.svxlog, name="svxlog"), path('survexfile/.err', survex.err, name="err"), path('survexfile/', survex.survexcavesingle, name="survexcavessingle"), path('survexfilewild', statistics.svxfilewild, name="svxfilewild"), path('survexfilewild/', statistics.svxfilewild, name="svxfilewild"), # The survey scans in the wallets. This short-cuts SCANS_URL which is not used anymore and is defunct path('survey_scans', allscans, name="allscans"), # all the scans in all wallets path('survey_scans//', walletedit, name="singlewallet"), # replaced singlewallet() path('survey_scans//', scansingle, name="scansingle"), # works, but html href goes direct to /expofiles/ too path('cave/scans/', cavewallets, name="cavewallets"), # like allscans, but for just one cave # The data about the wallets themselves, not the scans inside them path('wallets/year/', walletslistyear, name="walletslistyear"), # wallets that are for a specific year, as an integer '1985' path('wallets/person/', walletslistperson, name="walletslistperson"), # The tunnel and therion drawings files pageswalletslistcave path('drawings', dwgallfiles, name="dwgallfiles"), path('dwgfiles', dwgallfiles, name="dwgallfiles"), path('dwgdataraw/', dwgfilesingle, name="dwgfilesingle"), # QMs pages - must precede other /caves pages? re_path(r'^cave/qms/([^/]+)/?$', caveQMs, name="caveQMs"), re_path(r'^cave/openqms/([^/]+)/?$', caveQMs, {'open': True}, name="cave_openQMs"), re_path(r'^cave/qms/(?P[^/]+)/(?P\d\d\d\d)-(?P\d*)(?P[ABCDXV\?]?)-?(?P[a-zA-Z]+.*)?$', qm, name="qm"), # Dogs breakfast # the resolution of a QM uses several fields together, there is no clean slug field. Artefact of history. # Prospecting Guide document re_path(r'^prospecting_guide/$', prospecting), # disabled. Bad links, incompatible image package use and very, very out of date. # This next set are all intercepted by Apache, if it is running, with no problem. re_path(r'^photos/(?P.*)$', mediapage, {'doc_root': settings.PHOTOS_ROOT}, name="mediapage"), # photo galleries re_path(r'^map/slippy/map.html', map, name="map"), # Redirects to OpenStreetMap JavaScript. NOT WORKING re_path(r'^map/(?P.*)$', mapfile, name="mapfile"), # css, js, gpx. working. This does not add any troggle menus or headers # This next set are all intercepted by Apache, if it is running, AND troggle must manage these, # even though the code is not in the troggle repo # Alias /static/ /home/expo/static/ re_path(r'^static/(?P.*)$', mediapage, {'doc_root': settings.MEDIA_ROOT}, name="mediapage"), # STATIC only used by admin pages # Alias /javascript /usr/share/javascript path('javascript/', mediapage, {'doc_root': settings.JSLIB_ROOT}, name="mediapage"), # JSLIB_URL # Helpers to edit HTML re_path(r'^image_selector/(?P.*)', image_selector, name = 'image_selector'), re_path(r'^new_image_form/(?P.*)', new_image_form, name = 'new_image_form'), re_path(r'^expo.kml', expo_kml, name = 'expo.kml'), re_path(r'^expo.kmz', expo_kmz, name = 'expo.kmz'), # Final catchall which also serves expoweb handbook pages and imagestiny # but a universal catchall also prevents the djang standard append_slash working, as every string resolves. # try to fix in troggle/middleware.py re_path(r'^(.*)$', expopage, name="expopage"), # CATCHALL assumed relative to EXPOWEB. This means APPEND_SLASH never works. ] # do NOT allow DIR_ROOT prefix to all urls #urlpatterns = static(settings.JSLIB_URL, document_root=settings.JSLIB_ROOT) #For development purposes, in production this should be served statically perhaps from /usr/share/javascript #urlpatterns += static(settings.EXPOWEBCACHE_URL, document_root=settings.EXPOWEBCACHE_ROOT) #For development purposes, in production this should be served statically urlpatterns = [re_path('', include(trogglepatterns))] # When apache is running these prempt Django so Django never sees them. # NB apache has its own ideas about mimetypes, so behaviour may not be identical for .xml files by troggle # These directives are in /home/expo/config/apache/expo.conf on the server # checked as correct 3 May 2023. #These are the critical ones: /static/ and /javascript/ #Note that apache does NOT intercept /site-media/ # Alias /static/ /home/expo/static/ # Alias /javascript /usr/share/javascript # # Alias /expofiles /home/expo/expofiles # Alias /photos /home/expo/webphotos # Alias /map /home/expo/expoweb/map # Alias /expowebcache /home/expo/expowebcache/ # no longer used # Alias /robots.txt /home/expo/static/robots.txt # Alias /favicon.ico /home/expo/static/favicon.ico #This is archaic ? # Redirect permanent "/expoimages/" "/expofiles/" # #kanboard, the trello-clone, but need individual user id to be effective # Alias /kanboard /home/expo/kanboard # #Search function (xapian-omega) # ScriptAlias /search /usr/lib/cgi-bin/omega/omega # ScriptAlias /hgrepositories /home/expo/config/apache/services/hgweb/hgweb.cgi # #bank of expo # #current expedition # ScriptAlias "/boe" "/home/expo/boe/boc/boc.pl" # #archived expedition boe instances # ScriptAliasMatch "^/boe-(.*)" "/home/expo/boe/boc-$1/boc.pl"