troggle/core/views/uploads.py

import subprocess
import hashlib
from datetime import datetime
from pathlib import Path

from django import forms
from django.core.files.storage import FileSystemStorage
from django.shortcuts import render, redirect

import settings
from troggle.core.models.logbooks import LogbookEntry, PersonLogEntry
from troggle.core.models.survex import DrawingFile

# from databaseReset import reinit_db # don't do this. databaseRest runs code *at import time*

from .auth import login_required_if_public

"""File upload 'views' 
Note that there are other file upload forms in views/wallet_edit.py
and that core/forms.py contains Django class-based forms for caves and entrances.
"""

todo = """
- munge the URL of images in the logbook entry so that they work from both the " /logbookedit/" page,
  the logbook /years/2023/ page and the logbook fragment page /logbookentry/<date>/
  Note that this munging has already been done when the entry is imported into the database, so 
  when doing an online edit it has already been fixed.

- Ideally we should validate uploaded file as being a valid file type, not a dubious script or hack
  Validate image files using a magic recogniser in walletedit()
  https://pypi.org/project/reportlab/   or 
  https://stackoverflow.com/questions/889333/how-to-check-if-a-file-is-a-valid-image-file

- Write equivalent GPX upload form system, similar to walletedit() but in expofiles/gpslogs/
  Need to validate it as being a valid GPX file using an XML parser, not a dubious script or hack
  
- Validate Tunnel & Therion files using an XML parser in dwgupload(). Though Julian says 
  tunnel is  only mostly correct XML, and it does fail at least one XML parser.
  
- parse the uploaded drawing file for links to wallets and scan files as done 
  in parsers/drawings.py

- Enable folder creation in dwguploads or as a separate form

- Enable file rename on expofiles, not just for /surveyscans/ (aka wallets)

- Make file rename utility less ugly.
"""
sha = hashlib.new('sha256')


class FilesForm(forms.Form):  # not a model-form, just a form-form
    uploadfiles = forms.FileField()

class FilesRenameForm(forms.Form):  # not a model-form, just a form-form
    uploadfiles = forms.FileField()
    renameto = forms.CharField(strip=True, required=False)

class TextForm(forms.Form):  # not a model-form, just a form-form
    photographer = forms.CharField(strip=True)
    
class ExpofileRenameForm(forms.Form):  # not a model-form, just a form-form
    renameto = forms.CharField(strip=True, required=False)
    
class LogbookEditForm(forms.Form):  # not a model-form, just a form-form
    author = forms.CharField(strip=True, required=False)
    
@login_required_if_public
def logbookedit(request, year=None, slug=None):
    """Edit a logbook entry
    This is daft: we have the parsed identity of the person and we render it to text as 'nickname_used' 
    (or, previously, 'fullname'), to be re-parsed on re-importing.
    And there is no guarantee that this will be the same thing.
    
    Someone can put in a nickname which is invalid (e.g. 2 Sophies on expo). When is this checked?
    """
    def clean_tu(tu):
        if tu =="":
            return 0
        try:
            tu = float(tu)/1 # check numeric
        except:
            return 0
        return tu
        
    def unique_id(text, n):
        """This gives each logbook entry a unique id based on the date+content, so the order of entries on a particular day
        does not matter. This is a change (August 2023) from previous process.
        Otherwise we could get 2023-07-20a and 2023-07-20b swapped on exporting and re-importing logbooks
        because the database does not record precedence.
        2 hex digits would seem adequate for each expo day, but we might get a collision.
        The hash is based on the content after substitution of <p> so should be stable. Which means these ids
        can be used elsewhere in the troggle system as permanent slugs.
        
        When SAVING an edited entry (as opposed to a new one) we will have a different hash so we will have to 
        delete the original database object
        """
        sha.update(text.encode('utf-8'))
        return sha.hexdigest()[0:n]
    
    if not year:
        if not slug:
            year = 2023
        else:
            year = slug[0:4]
            print(year)
    
    if request.method == "POST":
        form = LogbookEditForm(request.POST)
        if not form.is_valid():
            message = f'Invalid form response for logbook entry creating "{request.POST}"'
            print(message)
            return render(request, "errors/generic.html", {"message": message})
        else:
            # validation all to be done yet..
            date = request.POST["date"].strip()  
            author = request.POST["author"].strip()   # TODO check against personexpedition
            others = request.POST["others"].strip()    # TODO check each against personexpedition
            place = request.POST["place"].strip().replace('-','=')  # no hyphens !
            title = request.POST["title"].strip()   
            entry = request.POST["text"].strip() 
            entry = entry.replace('\r','') # remove HTML-standard CR inserted
            entry = entry.replace('\n\n','\n<br />\n<br />\n') # replace 2 \n  with <br><br>
            entry = entry.replace('<p','<br />\n<br') # replace <p> tag, even if it has attributes,  with <br><br>
            entry = entry.replace('<br>','<br />') # clean up previous hack
            tu = request.POST["tu"].strip()
            tu = clean_tu(tu)
            
            try:
                odate = datetime.strptime(date.replace(".", "-"), "%Y-%m-%d").date()
                dateflag = False
            except:
                odate = datetime.strptime(f"{year}-01-01", "%Y-%m-%d").date()
                print(f"! Invalid date string {date}, setting to {odate}")
                dateflag = True
            date = odate.isoformat()
            
            newslug = f"{date}_{unique_id(entry,2)}"
            if slug:
                if slug != newslug:
                    print(f"! Entry id changed! from {slug} to {newslug}")
            
            # OK this could be done by rendering a template, but for such a small bit of HTML, it is easier to have
            # it all in one place: here
            output = f'''

<div class="tripdate" id="{newslug}">{date}</div>
<div class="trippeople"><u>{author}</u>, {others}</div>
<div class="triptitle">{place} - {title}</div>

{entry}

<div class="timeug">T/U {tu} hrs</div>
<hr />

'''
            return render(
                request,
                "logbookform.html",
                {
                    "form": form,
                    "year": year,
                    "date": date, "dateflag": dateflag,
                    "author": author,
                    "others": others,
                    "place": place,
                    "title": title,
                    "tu": tu,
                    "entry": entry,
                    "output": output,
                },
            )
    # GET here
    else:
        form = LogbookEditForm()
        
        if slug:
            lbes = LogbookEntry.objects.filter(slug=slug)
            if lbes:
                if len(lbes) > 1:
                     return render(request, "object_list.html", {"object_list": lbe}) # ie a bug
                else:
                    lbe = lbes[0]
                print(f"{lbe}")
                tu = clean_tu(lbe.time_underground)
                
                people = []
                for p in lbe.personlogentry_set.filter(logbook_entry=lbe): # p is a PersonLogEntry object
                    if p.is_logbook_entry_author:
                        # author = p.personexpedition.person.fullname
                        author = p.nickname_used
                    else:
                        # people.append(p.personexpedition.person.fullname)
                        people.append(p.nickname_used)
                others =', '.join(people)
                lenothers = min(70,max(20, len(others)))
                print(f"{lenothers}")
                text = lbe.text
                rows = max(5,len(text)/50)
                return render(
                    request,
                    "logbookform.html",
                    {
                        "form": form,
                        "year": year,
                        "date": lbe.date.isoformat(),
                        "author": author,
                        "others": others,
                        "lenothers": lenothers,
                        "place": lbe.place,
                        "title": lbe.title.replace(f"{lbe.place} - ",""),
                        "tu": tu,
                        "entry": text,
                        "textrows": rows,
                        #"output": output,
                    },
                )
        else:
            return render(
                request,
                "logbookform.html",
                {
                    "form": form,
                    "year": year,
                },
            )



    
@login_required_if_public
def expofilerename(request, filepath):
    """Rename any single file in /expofiles/ - eventually.
    Currently this just does files within wallets i.e. in /surveyscans/
    and it returns control to the original wallet edit page
    """
    if filepath:
        actualpath = Path(settings.EXPOFILES) / Path(filepath)
    else:
        message = f'\n File to rename not specified "{filepath}"'
        print(message)
        return render(request, "errors/generic.html", {"message": message})
    
    if not actualpath.is_file():
        message = f'\n File not found when attempting rename "{filepath}"'
        print(message)
        return render(request, "errors/generic.html", {"message": message})
    else:
        filename = Path(filepath).name
        folder = actualpath.parent
        filesize = f"{actualpath.stat().st_size:,}"

    if not actualpath.is_relative_to(Path(settings.SCANS_ROOT)):
        message = f'\n Can only do rename within wallets (expofiles/surveyscans/) currently, sorry. "{actualpath}" '
        print(message)
        return render(request, "errors/generic.html", {"message": message})
     
    if request.method == "POST":
        form = ExpofileRenameForm(request.POST)
        if not form.is_valid():
            message = f'Invalid form response for file renaming "{request.POST}"'
            print(message)
            return render(request, "errors/generic.html", {"message": message})
        else:
            renameto = request.POST["renameto"]
            
            if (folder / renameto).is_file() or (folder / renameto).is_dir():
                rename_bad = renameto
                message = f'\n Cannot rename to an existing file or folder. "{filename}" -> "{(folder / renameto)}"'
                print(message)
                return render(
                    request,
                    "renameform.html",
                    {
                        "form": form,
                        "filepath": filepath,
                        "filename": filename,
                        "filesize": filesize,
                        "rename_bad": rename_bad,
                    },
                )
            else:
                actualpath.rename((folder / renameto))
                message = f'\n RENAMED "{filename}" -> "{(folder / renameto)}"'
                print(message)
                walletid  = actualpath.relative_to(Path(settings.SCANS_ROOT)).parent.stem.replace("#",":")
                print(walletid)
                return redirect(f'/survey_scans/{walletid}/')
                
    else:
        form = ExpofileRenameForm()
        return render(
            request,
            "renameform.html",
            {
                "form": form,
                "filepath": filepath,
                "filename": filename,
                "filesize": filesize,
            },
        )
 
@login_required_if_public
def photoupload(request, folder=None):
    """Upload photo image files into /expofiles/photos/<year>/<photographer>/
    This does NOT use a Django model linked to a Django form. Just a simple Django form.
    You will find the Django documentation on forms very confusing, This is simpler.


    When uploading from a phone, it is useful to be able to rename the file to something
    meaningful as this is difficult to do on a phone. Previously we had assumed files would
    be renamed to something useful before starting the upload.
    Unfortunately this only works when uploading one file at at time ,
    inevitable once you think about it.
    
    Pending generic file renaming capability more generally.
    """
    year = settings.PHOTOS_YEAR
    filesaved = False
    actual_saved = []

    context = {"year": year, "placeholder": "AnathemaDevice"}

    yearpath = Path(settings.PHOTOS_ROOT, year)

    if folder == str(year) or folder == str(year) + "/":
        folder = None

    if folder is None:
        folder = ""  # improve this later
        dirpath = Path(settings.PHOTOS_ROOT, year)
        urlfile = f"/expofiles/photos/{year}"
        urldir = f"/photoupload/{year}"
    else:  # it will contain the year as well as the photographer
        dirpath = Path(settings.PHOTOS_ROOT, folder)
        if dirpath.is_dir():
            urlfile = f"/expofiles/photos/{folder}"
            urldir = Path("/photoupload") / folder
        else:
            folder = ""  # improve this later
            dirpath = Path(settings.PHOTOS_ROOT, year)
            urlfile = f"/expofiles/photos/{year}"
            urldir = f"/photoupload/{year}"

    form = FilesRenameForm()
    formd = TextForm()

    if request.method == "POST":
        if "photographer" in request.POST:
            formd = TextForm(request.POST)
            if formd.is_valid():
                newphotographer = request.POST["photographer"]
                try:
                    (yearpath / newphotographer).mkdir(exist_ok=True)
                except:
                    message = f'\n !! Permissions failure ?! 0 attempting to mkdir "{(yearpath / newphotographer)}"'
                    print(message)
                    return render(request, "errors/generic.html", {"message": message})

        else:
            form = FilesRenameForm(request.POST, request.FILES)
            if form.is_valid():
                f = request.FILES["uploadfiles"]
                multiple = request.FILES.getlist("uploadfiles")
                # NO CHECK that the files being uploaded are image files
                fs = FileSystemStorage(dirpath)

                renameto = request.POST["renameto"]

                actual_saved = []
                if multiple:
                    if len(multiple) == 1:
                        if renameto != "":
                            try:  # crashes in Django os.chmod call if on WSL, but does save file!
                                saved_filename = fs.save(renameto, content=f)
                            except:
                                print(
                                    f'\n !! Permissions failure ?! 1 attempting to save "{f.name}" in "{dirpath}" {renameto=}'
                                )
                                if "saved_filename" in locals():
                                    if saved_filename.is_file():
                                        actual_saved.append(saved_filename)
                                        filesaved = True
                        else:  # multiple is the uploaded content
                            try:  # crashes in Django os.chmod call if on WSL, but does save file!
                                saved_filename = fs.save(f.name, content=f)
                            except:
                                print(
                                    f'\n !! Permissions failure ?! 2 attempting to save "{f.name}" in "{dirpath}" {renameto=}'
                                )
                                if "saved_filename" in locals():
                                    if saved_filename.is_file():
                                        actual_saved.append(saved_filename)
                                        filesaved = True
                    else:  # multiple is a list of content
                        for f in multiple:
                            try:  # crashes in Django os.chmod call if on WSL, but does save file!
                                saved_filename = fs.save(f.name, content=f)
                            except:
                                print(
                                    f'\n !! Permissions failure ?! 3 attempting to save "{f.name}" in "{dirpath}" {renameto=}'
                                )
                                if "saved_filename" in locals():
                                    if saved_filename.is_file():
                                        actual_saved.append(saved_filename)
                                        filesaved = True
    files = []
    dirs = []
    try:
        for f in dirpath.iterdir():
            if f.is_dir():
                dirs.append(f.name)
            if f.is_file():
                files.append(f.name)
    except FileNotFoundError:
        files.append("(no folder yet - would be created)")
    if len(files) > 0:
        files = sorted(files)

    if dirs:
        dirs = sorted(dirs)

    return render(
        request,
        "photouploadform.html",
        {
            "form": form,
            **context,
            "urlfile": urlfile,
            "urldir": urldir,
            "folder": folder,
            "files": files,
            "dirs": dirs,
            "filesaved": filesaved,
            "actual_saved": actual_saved,
        },
    )


@login_required_if_public
def dwgupload(request, folder=None, gitdisable="no"):
    """Upload DRAWING files (tunnel or therion) into the upload folder in :drawings
    AND registers it into the :drawings: git repo.

    This does NOT use a Django model linked to a Django form. Just a simple Django form.
    You will find the Django documentation on forms very confusing, This is simpler.

    We could validate the uploaded files as being a valid files using an XML parser, not a dubious script or hack,
    but this won't work on Tunnel files as Tunnel does not produce exactly valid xml

    We use get_or_create instead of simply creating a new object in case someone uploads the same file
    several times in one session, and expects them to be overwritten in the database. Although
    the actual file will be duplicated in the filesystem with different random name ending.
    """

    def dwgvalid(name):
        if name in [
            ".gitignore",
        ]:
            return False
        if Path(name).suffix.lower() in [".xml", ".th", ".th2", "", ".svg", ".txt"]:
            return True  # dangerous, we should check the actual file binary signature
        return False

    def dwgvaliddisp(name):
        """OK to display, even if we are not going to allow a new one to be uploaded"""
        if name in [
            ".gitignore",
        ]:
            return False
        if Path(name).suffix.lower() in [
            ".xml",
            ".th",
            ".th2",
            "",
            ".svg",
            ".txt",
            ".jpg",
            ".jpeg",
            ".png",
            ".pdf",
            ".top",
            ".topo",
        ]:
            return True  # dangerous, we should check the actual file binary signature
        return False

    filesaved = False
    actual_saved = []
    refused = []
    doesnotexist = ""
    # print(f'! - FORM dwgupload - start "{folder}" - gitdisable "{gitdisable}"')
    if folder is None:
        folder = ""  # improve this later
        dirpath = Path(settings.DRAWINGS_DATA)
        urlfile = "/dwgdataraw"
        urldir = "/dwgupload"
    else:
        dirpath = Path(settings.DRAWINGS_DATA, folder)
        urlfile = Path("/dwgdataraw/") / folder
        urldir = Path("/dwgupload/") / folder

    form = FilesForm()

    if request.method == "POST":
        form = FilesForm(request.POST, request.FILES)
        if form.is_valid():
            # print(f'! - FORM dwgupload - POST valid: "{request.FILES["uploadfiles"]}" ')
            f = request.FILES["uploadfiles"]
            multiple = request.FILES.getlist("uploadfiles")
            savepath = Path(settings.DRAWINGS_DATA, folder)
            fs = FileSystemStorage(savepath)

            actual_saved = []
            refused = []

            # GIT see also core/views/expo.py editexpopage()
            # GIT see also core/models/cave.py writetrogglefile()
            if gitdisable != "yes":  # set in url 'dwguploadnogit/'
                git = settings.GIT
            else:
                git = "echo"
                # print(f'git DISABLED {f.name}')

            if multiple:
                for f in multiple:
                    # print(f'! - FORM dwgupload - file {f} in {multiple=}')
                    if dwgvalid(f.name):
                        try:  # crashes in Django os.chmod call if on WSL without metadata drvfs, but does save file!
                            saved_filename = fs.save(f.name, content=f)
                        except:
                            print(
                                f'! - FORM dwgupload - \n!! Permissions failure ?! on attempting to save file "{f.name}" in "{savepath}". Attempting to continue..'
                            )
                        if "saved_filename" in locals():
                            if Path(dirpath, saved_filename).is_file():
                                actual_saved.append(saved_filename)
                                if gitdisable != "yes":
                                    dr_add = subprocess.run(
                                        [git, "add", saved_filename], cwd=dirpath, capture_output=True, text=True
                                    )
                                    msgdata = (
                                        dr_add.stderr
                                        + "\n"
                                        + dr_add.stdout
                                        + "\nreturn code: "
                                        + str(dr_add.returncode)
                                    )
                                    # message = f'! - FORM dwgupload - Success: git ADD on server for this file {saved_filename}.' + msgdata
                                    # print(message)
                                    if dr_add.returncode != 0:
                                        msgdata = (
                                            "Ask a nerd to fix this.\n\n"
                                            + dr_add.stderr
                                            + "\n\n"
                                            + dr_add.stdout
                                            + "\n\nreturn code: "
                                            + str(dr_add.returncode)
                                        )
                                        message = (
                                            f"! - FORM dwgupload - CANNOT git ADD on server for this file {saved_filename}. Edits saved but not added to git.\n"
                                            + msgdata
                                        )
                                        print(message)
                                        return render(request, "errors/generic.html", {"message": message})
                                dwgfile, created = DrawingFile.objects.get_or_create(
                                    dwgpath=saved_filename, dwgname=Path(f.name).stem, filesize=f.size
                                )
                                dwgfile.save()
                            else:
                                message = f"! - FORM dwgupload - NOT A FILE {Path(dirpath, saved_filename)=}. "
                                print(message)
                        else:
                            message = f"! - FORM dwgupload - Save failure for {f.name}. Changes NOT saved."
                            print(message)
                            return render(request, "errors/generic.html", {"message": message})

                        if saved_filename != f.name:
                            # message = f'! - FORM dwgupload - Save RENAME {f.name} renamed as {saved_filename}. This is OK.'
                            # print(message)
                            pass

                    else:
                        refused.append(f.name)
                        # print(f'REFUSED {f.name}')

            if actual_saved:
                filesaved = True
                if len(actual_saved) > 1:
                    dots = "..."
                else:
                    dots = ""
                if gitdisable != "yes":
                    dr_commit = subprocess.run(
                        [git, "commit", "-m", f"Drawings upload - {actual_saved[0]}{dots}"],
                        cwd=dirpath,
                        capture_output=True,
                        text=True,
                    )
                    # message = f'! - FORM dwgupload - For uploading {actual_saved[0]}{dots}. Edits saved, added to git, and COMMITTED.\n' + msgdata
                    # print(message)
                    # This produces return code = 1 if it commits OK
                    if dr_commit.returncode != 0:
                        msgdata = (
                            "Ask a nerd to fix this.\n\n"
                            + dr_commit.stderr
                            + "\n"
                            + dr_commit.stdout
                            + "\nreturn code: "
                            + str(dr_commit.returncode)
                        )
                        message = (
                            f"! - FORM dwgupload -Error code with git on server for this {actual_saved[0]}{dots}. Edits saved, added to git, but NOT committed.\n"
                            + msgdata
                        )
                        print(message)
                        return render(request, "errors/generic.html", {"message": message})
                else:
                    print(f' git disabled "{git=}"')
            else:  # maybe all were refused by the suffix test in dwgvalid()
                message = f"! - FORM dwgupload - Nothing actually saved. All were refused. {actual_saved=}"
                print(message)

    files = []
    dirs = []
    # print(f'! - FORM dwgupload - start {folder=} \n"{dirpath=}" \n"{dirpath.parent=}" \n"{dirpath.exists()=}"')
    try:
        for f in dirpath.iterdir():
            if f.is_dir():
                if f.name not in [".git"]:
                    dirs.append(f.name)
                    continue
            if f.is_file():
                if dwgvaliddisp(f.name):
                    files.append(f.name)
                    continue
    except FileNotFoundError:
        doesnotexist = True
    if files:
        files = sorted(files)

    if dirs:
        dirs = sorted(dirs)

    return render(
        request,
        "dwguploadform.html",
        {
            "form": form,
            "doesnotexist": doesnotexist,
            "urlfile": urlfile,
            "urldir": urldir,
            "folder": folder,
            "files": files,
            "dirs": dirs,
            "filesaved": filesaved,
            "actual_saved": actual_saved,
            "refused": refused,
        },
    )