Egypt (EG) — internal licensing notes

Source research, 2026-09-11 — tier-1 check done, real official source found but not bulk-fetchable

Egypt has a real, structured 7-digit postal code system — confirmed directly from UPU's own addressing-format reference for Egypt (upu.int/.../addressingUnit/egyEn.pdf, fetched and read in full 2026-09-11). The 7 digits decode as province + locality + neighbourhood + community (e.g. 3759914 = Giza province code + locality + neighbourhood

• community digits) — a real hierarchical system, not an arbitrary number, covering roughly 8,500 post office locations per independent reporting (not verified directly here).

That UPU document is a format reference, not a data source — it describes the coding scheme with worked examples, it does not contain or link to a bulk list.

The official source — Egypt Post

egyptpost.org redirects to egyptpost.gov.eg — blocked (403 Forbidden) on direct fetch from this environment; could not confirm what tools it offers beyond what search results describe (a postcode-lookup page at egyptpost.gov.eg/ar-eg/home/eservices/find-postal-code/). No bulk download, open-data catalogue entry (data.gov.eg), or API was found for Egypt's postal codes in this pass.

Why GeoNames is not a fallback here

download.geonames.org/export/zip/EG.zip returns a plain 404 — no GeoNames coverage exists for Egypt at all.

Verdict

No build in this pass. A real, structured official system exists and its format is well-documented, but no bulk-fetchable source was found — Egypt Post's own lookup page could not be directly inspected (blocked), and no open-data portal entry surfaced. data_available unchanged (false). Worth a follow-up from an environment that isn't blocked by Egypt Post's site, and a direct check of data.gov.eg's catalogue rather than search results.

Re-check, 2026-09-15 — fresh search, conclusion unchanged

Re-ran source discovery from scratch (not just re-reading the above) per a request to conclusively confirm or overturn the 2026-09-11 verdict. Egypt is explicitly NOT in the same category as no-postal-code countries (Bolivia/Ethiopia/Uganda/Libya/Zimbabwe) built as free locality directories in this repo — it has a real, documented 7-digit hierarchical postal-code system, so a locality-only substitute would misrepresent the product and is out of scope regardless of what this re-check finds (same reasoning as MT and BF, both re-checked this same pass). Checked, this pass:

1. Egypt Post, retried with a different approach. The 2026-09-11 note only recorded a blocked fetch without detail on method. This pass retried directly with curl using a full browser User-Agent string (Mozilla/5.0 ... Chrome/120.0) against both https://www.egyptpost.gov.eg/ and https://egyptpost.gov.eg/ar-eg/home/eservices/find-postal-code/ — both still return a hard 403 Forbidden. Separately tried WebFetch against the postcode-finder URL — also 403 Forbidden. Two independent tools, two different request signatures, same block. This is a genuine, confirmed block, not a one-off fluke — no further variations attempted (would tip into tunnel-vision on a single blocked domain, which the task explicitly warned against).
2. data.gov.eg — does not resolve at all. curl -v against both https://data.gov.eg/ and the plain-HTTP form timed out at DNS resolution (Resolving timed out after 10001 milliseconds) — this is a dead domain, not a live portal, matching the same dead-domain pattern found for data.gov.bf in this session's Burkina Faso re-check.
3. egypt.gov.eg's "Open Gov Data Initiative" page — TLS certificate is expired (certificate has expired on fetch); forcing the connection through anyway returns a bare 302 redirect with no usable content. Effectively an abandoned/unmaintained government page, not a working catalogue.
4. CAPMAS (capmas.gov.eg / censusinfo.capmas.gov.eg) — re-checked. CAPMAS does publish administrative-boundary data (COD-AB, admin levels 0–3, also mirrored on HDX) and 2017 census locality files down to Shyakha/Qism level — but this is boundary/population geography, not postal codes; no postcode field exists in any CAPMAS product found. Irrelevant to this task twice over: it wouldn't answer the postal-code question even if usable, and a locality-only build is explicitly ruled out for EG regardless.
5. UPU Universal POST*CODE® Database — the free Africa sample was downloaded and directly inspected, and Egypt's file IS present (unlike Burkina Faso, where the equivalent check found zero rows). Fetched https://www.upu.int/UPU/media/upu/documents/PostCode/Africa.zip (v2026.1, dated 2026-03-17) directly — 840KB zip, 5 country folders: NGA, EGY, MAR, ZAF, SOM. The EGY/ folder contains two files:
   • EGY_001.dat (LOCALITY table): only 10 rows, all coded EG.QN — i.e. all 10 localities belong to a single governorate, Qina (قنا), one of Egypt's 27 governorates. Real place names present (Qena, Nagaa Hammadi, Abu Tisht, Nuqaadih, Qift, Quos, Dashna, Farshout, Alwaqf).
     • EGY_004.dat (ORGANIZATION table): 23,328 data rows, every one of them keyed back to one of those same 10 Qina localities (verified directly — the locality-ID column has exactly 11 distinct values including the header). Each row carries a genuine-looking 7-digit code (e.g. 1450224) consistent with the UPU addressing-format PDF read in the 2026-09-11 pass — this is real postal-code-shaped data, not a placeholder or dummy file.
   • This is real data, but it covers 1 of Egypt's 27 governorates (~3.7%) and nothing else — no Cairo, no Alexandria, no Giza, no Delta or Upper Egypt governorates beyond Qina. Measured the same way this project measured Malta's truncated GeoNames export (a ~99.7% coverage loss disqualified it as a fallback), this sample represents roughly a 96%+ coverage loss against the full country — nowhere near the "sufficiently complete" bar for a build.
   • No licence file is bundled inside Africa.zip itself — checked directly, no readme/terms/licence file present alongside the .dat files. The UPU addressing-solutions page's own text around the sample links only says the samples exist so a prospect can "explore this unique professional tool" and invites contact for more samples — marketing-teaser framing, not a free-use grant.
   • The full database is confirmed commercial/NDA-gated, same as the BF finding: downloaded and inspected LicenceDocumentsPostCodeEn.zip from the same page directly — it contains a Basic Contract, a User NDA, General Conditions, an In-house Rates schedule, and Exhibits listing the countries/ territories provided and the database structure. There is no path to the other 26 governorates without entering that paid/NDA relationship.
   • Conclusion on this specific lead: real and non-trivial, but not sufficiently complete and not freely licensed — it does not clear the bar this project uses elsewhere (order-of-thousands-scale, nationally complete, freely or clearly licensed) and does not change the build verdict. Flagging it precisely here rather than folding it into "same as BF, nothing found," since the two countries' UPU samples are genuinely different in kind (BF: zero rows; EG: real but partial data) and a future pass should not have to re-derive this.
6. GitHub — Tech-Labs/egypt-governorates-and-cities-db is an admin-only governorates/cities database (no postal codes) and is irrelevant regardless, per the locality-only exclusion above. No dedicated, sufficiently complete Egypt postcode repo was found.
7. Aggregators (egpostal.com, getpostalcodes.com/egypt, en.56ok.com/zipcode_EG.html) — all one-postcode/one-city-at-a-time lookup tools, no bulk export, no API, no stated licence terms on any of them. Not pursued further, same treatment as equivalent aggregators flagged elsewhere in this project.
8. This repo's own datahub.io/logistics/postal-codes-eg catalogue entry — re-checked directly: it explicitly states "Postal code data is sourced from official and publicly available records. No data is currently available for Egypt" and offers no download. Confirms no quiet fix has landed upstream since the last check.
9. GeoPostcodes — its own marketing figures (533,859 postcodes, 8,533 towns, 426 regions for Egypt) were the only claim of a complete commercial dataset found; same paid-competitor-product category noted elsewhere in this project (e.g. MT, BF) — not pursued, since buying or relicensing a competitor's commercial extract is out of scope for a source-discovery pass.

No new source clears the bar (bulk-downloadable, nationally-complete, freely or clearly licensed). The one genuinely new finding this pass — the UPU Africa sample actually containing real Egyptian postal data — does not change the outcome: it is real but covers only 1 of 27 governorates and carries no free-use licence, the same disqualifying shape as GeoNames' deliberately-truncated Malta export or GeoPostcodes' paid product. Conclusion from 2026-09-11 stands unchanged: do not build. data_available remains false. datasets/eg/datapackage.yml was not touched.

Note for a future pass: the only two real leads that would change this verdict are (a) Egypt Post agreeing to a direct bulk export or exposing an open-data catalogue (its own lookup site is confirmed blocked, not just undiscovered), or (b) budget/approval to license UPU's Universal POST*CODE® Database commercially for full 27-governorate coverage — a business decision outside a source-discovery pass, not a technical gap.