Using cursor.next() is resetting "row in cursor" causing it to skip every other row

Question

I'm trying to update a column based on the value in the next row, but cursor.next() is giving unexpected behavior. When I identify the next_row using cursor.next(), it is resetting my row in cursor! The result is my loop goes over every other row, not every row. What am I missing? (Code below is just to show cursor behavior; code for row update for column has been stripped.)

Code

import arcpy
temp1 = r'mypath\mypointfile.shp'
with arcpy.da.UpdateCursor(temp1, ['OID@']) as cursor:
    for row in cursor:
        try:
            row_next = cursor.next()
            print 'current_oid=', row[0], 'next_oid=', row_next[0]
        except StopIteration:
            print 'no rows'
        except:
            print 'something else went wrong'

Actual Prints

current_oid= 0 next_oid= 1
current_oid= 2 next_oid= 3
current_oid= 4 next_oid= 5
current_oid= 6 next_oid= 7
current_oid= 8 next_oid= 9

Expected Prints

current_oid= 0 next_oid= 1
current_oid= 1 next_oid= 2
current_oid= 2 next_oid= 3
current_oid= 3 next_oid= 4
current_oid= 4 next_oid= 5

UPDATE

Purpose of this is to only keep rows that are the first or last point in the group.

Example attributes:

oid group   id
0   1   1
1   1   2
2   1   3
3   2   1
4   2   2
5   3   1
6   3   2
7   3   3
8   3   4
9   3   5

Desired attributes:

group   id
1   1
1   3
2   1
2   2
3   1
3   5

Original code, that didn't work because cursor.next() skips rows:

with arcpy.da.UpdateCursor(temp1, ['OID@', 'group', 'id']) as cursor:
    for row in cursor:
        group_current = row[1]
        id_current = row[2]
        if id_current == 1:
            check = "keep"
        else:
            try:
                row_next = cursor.next()
                group_next = row_next[1]
                id_next = row_next[2]
                if id_next == 1:
                    check = "keep"
                else:
                    check = "remove"
            except StopIteration:
                check = "keep"
            except:
                print 'something else went wrong'
            if check != "keep":
                cursor.deleteRow()

Answer 1

Yes, this is expected as it's how python generators work, once a value is consumed, it's gone. If your feature class isn't toooo big, pull all the rows into a list, then use enumerate and indexing to access the next row. E.g.

cursor = ((i,) for i in range(10))  # Fake a cursor as I'm not at a PC with ArcGIS
rows = list(cursor)  # Exhaust the generator
for i, row in enumerate(rows):
    try:
        row_next = rows[i+1]
        print(row[0],row_next[0])
    except IndexError:
        print('no rows')

Output:

0 1
1 2
2 3
3 4
4 5
5 6
6 7
7 8
8 9
no rows

Or... turn it around, grab the first row before the loop then refer to the previous row instead of the next row:

prev_row = next(cursor)
for row in cursor:
    print(prev_row[0], row[0])
    prev_row = row

Output:

0 1
1 2
2 3
3 4
4 5
5 6
6 7
7 8
8 9

Answer 2

Thanks to Vince's comments and user2856's answer/comments, the following worked. Elapsed time = 30.17 seconds for 52,000 row, which seems slow.

temp1 = r'mypath\mypointfile.shp'
import datetime
now = datetime.datetime.now()

cursor = arcpy.da.SearchCursor(temp1, ['OID@', 'group', 'id'])
rows = list(cursor)
dict_ids = {}
for i, row in enumerate(rows):
    oid = row[0]
    group_current = row[1]
    id_current = row[2]
    if id_current == 1:
        check = "keep"
    else:
        try:
            row_next = rows[i + 1]
            id_next = row_next[2]
            if vid_next == 1:
                check = "keep"
            else:
                check = "remove"
        except IndexError:
            check = "keep"
        except:
            print 'something else went wrong'
    dict_ids.update([(oid, check)])
del i, row, check
with arcpy.da.UpdateCursor(temp1, ['OID@']) as cursor:
    for row in cursor:
        oid = row[0]
        if oid in dict_ids:
            check = dict_ids.get(oid)
        else:
            print 'error with dictionary'
            sys.exit()
        if check != "keep":
            cursor.deleteRow()

later = datetime.datetime.now()
elapsed = later - now
print elapsed