Fix typos and inconsistencies for HEAD (take 5)

Hello hackers,

Please consider fixing the next cluster of typos and inconsistencies in
the tree:
5.1. datetkntbl -> datetktbl
5.2. datminxmid -> datminmxid
5.3. DatumGetP -> DatumGetPointer
5.4. ECPG_COMPILE, DECPG_COMPILE -> remove (orphaned since 328d235)
5.5. defer_cleanup_age -> vacuum_defer_cleanup_age
5.6. descriptor_index -> remove (orphaned since 991b974)
5.7. DestroyBuilder, InitBuilder, SetDoc -> DestroyOpaque, InitOpaque,
SetDocument
5.8. dictlexize -> thesaurus_lexize
5.9. regression.diffsregression.planregress/inh -> regression.diffs
planregress/diffs.inh
5.10. dllist -> dlist
5.11. DocRepresentaion -> DocRepresentation
5.12. dosplit -> remove (such function is not present since introduction
in 9892ddf)
5.13. DOWN_MEM_FENCE -> _DOWN_MEM_FENCE
5.14. dp_pg_stop_backup -> do_pg_stop_backup
5.15. DropRelFileNodeAllBuffers -> DropRelFileNodesAllBuffers
5.16. dshash_release -> dshash_release_lock
5.17. EACCESS -> remove (not used since introduction in 12c94238)
5.18. ECPGcheck_PQresult -> ecpg_check_PQresult (renamed in 7793c6ec,
referenced code moved in ecpg_execute() in 61bee9f7)
5.19. ecpg_compatlib -> libecpg_compat
5.20. ECPGerrmsg -> remove (not used since introduction in 244d2d67)
5.21. ecpg_free_auto_mem -> remove (not used since introduction in 7793c6ec)
5.22. ecpggetdescp -> remove (not used since introduction in 90326c01)
5.23. endBlk -> numBlks
5.24. endMemb, startMemb, endOff, startOff -> endTruncMemb,
startTruncMemb, endTruncOff, startTruncOff
5.25. EndPlan -> ExecEndPlan
5.26. EndResult -> ExprEndResult
5.27. equivalentOpersAfterPromotion -> remove (irrelevant since
8536c962, but the whole comments is too old to be informational too)
5.28. es_jit_combined_instr -> es_jit_worker_instr (renamed in c03c1449)
5.29. ExclusiveRowLock -> RowExclusiveLock
5.30. exdended -> extended (user-visible, I would fix it in
REL_12_STABLE too)
5.31. ExecBitmapHeapNext -> BitmapHeapNext
5.32. ExecBuildProjectInfo -> ExecBuildProjectionInfo
5.33. ExecDirection -> remove (this variable is not present since PG95-1_01)
5.34. ExecEndRecursiveUnionScan -> ExecEndRecursiveUnion
5.35. ExecGrantStmt -> ExecuteGrantStmt
5.36. ExecInitRecursiveUnionScan -> ExecInitRecursiveUnion
5.37. ExecSeqNext -> SeqNext
5.38. exec_statement_return -> exec_stmt_return
5.39. exec_subplan_get_plan -> remove (not used since 1cc29fe7)
5.40. ExecSubqueryNext -> SubqueryNext
5.41. ExecValuesNext -> ValuesNext
5.42. existing_oid -> existing_relation_id
5.43. exit_fatal -> fatal
5.44. expectedTLIs -> expectedTLEs
5.45. ExprEvalExpr -> ExecEvalExpr
5.46. exprhasexecparam -> remove (orphaned since 6630ccad)
5.47. ExprReadyExpr -> ExecReadyExpr
5.48. EXTENSION_REALLY_RETURN_NULL -> remove (the behaviour changed with
a7124870)

There are some other ancient comments like spotted in 5.27, e.g. for
textcat(), text_substr() in varlena.c...  It seems that they serve more 
historic than informational purposes today.

Best regards,
Alexander