Exporting DDP to multiple multi-page pdf files

Question

In ArcMap 10.4 I have an index layer in DDP enabled mxd. The index layer has 50 unique Name values (used as the PageName in DDP) but there are more rows with the same Name value providing different scales, spatial extents etc. I'm looking for an ArcPy script to export pdf files with all pages of the same Name.

Example:

Name   Scale
A      5000
A      600
B      25000
B      5000
B      2500
...

Result:

A.pdf (2 pages)
B.pdf (3 pages)
...

Would you be so kind to provide some hint? Apparently, I'm stuck in the process of correctly defining the 'for' loop.

So far I tried to use the modified code from Cindy Jayakumar https://gis.stackexchange.com/a/71702/93026, which was closest to what I found:

import arcpy
import os

mxd = arcpy.mapping.MapDocument("CURRENT")
output_folder = r"C:\temp"
ddp_lyr = mxd.dataDrivenPages.indexLayer
pdf_field = "Name"

#Get a list of unique values to group by
pdf_list = set([row[0] for row in arcpy.da.SearchCursor(ddp_lyr,pdf_field)])

for pdfName in pdf_list:
    arcpy.mapping.PDFDocumentCreate(os.path.join(output_folder,pdfName + "_inv.pdf"))
    for i in range(1, mxd.dataDrivenPages.pageCount + 1):
        mxd.dataDrivenPages.currentPageID = i
        cur = str(mxd.dataDrivenPages.pageRow.getValue(pdf_field))
        if cur == pdfName:
            tempPdf = os.path.join(output_folder,"temp_" + cur + str(i) + ".pdf")
            arcpy.mapping.ExportToPDF(mxd, tempPdf)
            finalPdf.appendPages(tempPdf)
            os.remove(tempPdf)
    finalPdf.saveAndClose()
    del finalPdf, tempPdf

The above code does what I need to some extent as from some reason does not export the pdf files for all unique values in the "Name" field. I was not able to identify why some have been exported and some not. I know there is much easier way to do this as 2 years ago I already had the code working (though not saving it).

Answer 1

At the end (and after some hours of debugging) I was able to create simple code which does the export.

I'll set another example:

Sites   River
A       Nile
A       Nile
A       Nile
B       Nile
B       Nile
C       Amazon
C       Amazon
...

Code

import arcpy

mxd = arcpy.mapping.MapDocument("CURRENT")
ddp = mxd.dataDrivenPages
indexLayer = ddp.indexLayer
output_folder = r"C:\temp"
select = set(row[0] for row in arcpy.da.SearchCursor(indexLayer, 'Sites'))

for site in select:
    where_clause = '"Sites" = ' + "'%s'" %site
    arcpy.SelectLayerByAttribute_management (indexLayer, "NEW_SELECTION", where_clause)
    riv = set(row[0] for row in arcpy.da.SearchCursor(indexLayer, 'River'))
    ddp.exportToPDF(r"C:\Temp\\" + str(site) + "_" + str(riv)[7:-3] + ".pdf", "SELECTED" )
del mxd

The stripping parameter str(riv)[7:-3] in the definition of the exported pdf name had to be added to delete the 'extra' characters "set(['*'])".

Before deletion:

A_set([u'Nile']).pdf
B_set([u'Nile']).pdf
C_set([u'Amazon']).pdf

After deletion:

A_Nile.pdf
B_Nile.pdf
C_Amazon.pdf

I don't know why variable riv returns this extended format, while variable site is pure string value, however, the script works fine now.