Modern Data Formats�Introduction

Jakub Klímek

This work is licensed under a Creative Commons Attribution 4.0 International License.

Some developers think mainly in terms of apps...

2

My computer

My app

My database

My app

My data

Data as part of an app

In public administration….

3

Computer

App

Database

App

Data

Give me the data, I want to build another app

No, you only paid for an app

Data as part of an app

App

Possible vendor-lock

Suddenly, the app needs to work with another app

4

His computer

His app

His database

His app

His data

Her computer

Her app

Her database

Her app

Her data

I will send you that

I will send you this

Suddenly, the app needs to work with another app

5

I will send you that

I will send you this

Data independent of applications

6

My computer

My app

My database

My app

Data

My computer

My app

My database

My app

Data independent of applications

7

Data

OK, data is independent. More problems?

Use of improper formats for a given use case

• e.g. tabular data in hierarchical data format (XML, JSON)

Not following specifications

• errors or unnecessary work when data is shared
• whose fault is it?
• the producer?
• the consumer?
• ultimately, who will pay for mitigating the issue?

8

Point of this course

1. overview of data formats, specifications, tools and use cases
2. ability to choose a proper data format for a given use case
3. thinking about data independently of applications and at various levels of abstraction
• conceptual level - what is the data about
• logical level - how is data structured using given technology/format
• physical level - how do the files look like in storage

9

Conceptual view of data

10

Conceptual domain model

Independent of any particular technology or representation. Answers:

What real world entities are described?

What are their properties?

How are they connected?

�Conceptual model can be discussed with non-IT personnel.

11

Example: National Open Data Catalog

12

Example: National Open Data Catalog - Dataset

13

Conceptual domain model - UML Class diagrams

Class: Catalog

��

This is saying:�“There are things in the real-world of a type Catalog.”

14

Conceptual domain model - UML Class diagrams

Class: Catalog�Attributes: title, description, homepage

​

This is saying:�“Each instance of a catalog has a title, description and a homepage.”

15

Conceptual domain model - UML Class diagrams

Class: Catalog�Attributes: title, description, homepage

Class: Contact point�Attributes: name, e-mail

​

This is saying:�“There are contact points, each has a name and an e-mail”

16

Conceptual domain model - UML Class diagrams

Class: Catalog�Attributes: title, description, homepage

Class: Contact point�Attributes: name, e-mail

Association: contact point

• Association end 1: Catalog, 0..*
• Association end 2: Contact point, 0..1

This is saying:

• An instance of a Catalog may or may not be connected to up to 1 instance of Contact point
• An instance of a Contact point may or may not be connected to an arbitrary number of instances of Catalog

17

Conceptual domain model - UML Class diagrams

Class: Dataset�Attributes: title, description

Association: dataset

• Association end 1: Catalog, 0..*
• Association end 2: Dataset, 0..*

This is saying:

• An instance of a Catalog may or may not be connected to an arbitrary number of instances of Dataset
• An instance of a Dataset may or may not be connected to an arbitrary number of instances of Catalog

18

Conceptual domain model - UML Class diagrams

Attributes of Dataset:

• keywords [0..*], specification [0..*]
• documentation
• spatial resolution, temporal resolution

Association: part of

• Association end 1: Dataset, 0..1
• Association end 2: Dataset, 0..*

Association: contact point

• Association end 1: Contact point, 0..1
• Association end 2: Dataset, 0..*

​

19

Conceptual domain model - UML Class diagrams

Associations:

• theme
• frequency
• item from RÚIAN
• spatial coverage
• EuroVoc classification
• Time interval

20

Conceptual domain model - UML Class diagrams

Association: distribution

• Association end 1: � Dataset, 1..1
• Association end 2: � Distribution, 0..*

21

Conceptual domain model - UML Class diagrams

Association: media type

• Association end 1: Distribution, 0..*
• Association end 2: Media type, 1..1

Association: package format

• Association end 1: Distribution, 0..*
• Association end 2: Media type, 0..1

Association: compression format

• Association end 1: Distribution, 0..*
• Association end 2: Media type, 0..1

22

Conceptual domain model - UML Class diagrams

23

Conceptual domain model - UML Class diagrams

24

Conceptual domain model - UML Class diagrams

DCAT-AP-CZ

• Czech specification of how to represent data catalogs
• 2021
• Based on DCAT-AP 2.1.1
• European application profile
• European Commission, 2022
• Based on DCAT 2
• W3C Recommendation, 2020

Representations in:

RDF, JSON, CSV

​

25

Data models

vs.�Data formats

vs.�Data schemas

26

Data models - logical view of data - graphs

Resource Description Framework (RDF) model

Labeled Property Graph (LPG) model

27

https://...

https://...

https://...

https://...

"Coffee shop"

Ann

Dan

Place

V60

Steve

John

KNOWS

KNOWS

EMPLOYS

VISITS

SERVES

EMPLOYS

:Person

:Person

:Person

:Person

:CoffeeShop

:BrewingMethod

name: V60

duration: 3 minutes

name: John�age: 42

name: Steve�age: 24

name: Dan�age: 87

name: Ann�age: 16

since: 2021-03-01

since: 2010-12-31

since: 2020-03-16

"Ann"

https://...

https://...

https://...

https://...

https://...

https://...

https://...

Data models - logical view of data - hierarchies/trees

Document Object Model (DOM)

JSON (both format and model)

28

document

root�<coffeeShops>

attribute�number

element�<coffeeShop>

element�<name>

element�<employees>

element�<employee>

text�"Place"

text�2

element�<name>

element�<age>

array

object

array

value�"Place"

object

object

name

employees

object

object

object

value�"Ann"

name

Data models - logical view of data - tables

Relational model

​

29

Data formats - physical view of data

Graph

• RDF graph model
• Text-based: N-Triples, N-Quads, Turtle, TriG, RDF/XML, JSON-LD, RDFa
• Binary: HDT��
• Property graph
• CSV, JSON, GraphML, Cypher Script

Hierarchical

• DOM
• XML, HTML
• JSON
• JSON, XML

30

How data using a certain data model is serialized into files / sent over network

Relational

• CSV, SQL dump

Data schemas

Annotations and constraints applicable to instances of data formats, allowing the data to be better described and validated

CSV

• Schema language
• CSV on the Web

RDF

• Schema language
• SHACL
• ShEx

JSON

• Schema language
• JSON Schema

XML

• Schema languages
• DTD
• XML Schema
• Relax NG
• Schematron

31

Specific data formats using meta-formats

CSV, JSON, XML, … sometimes called meta-formats

They serve as “host” formats for use-case specific formats

Data schemas used to define these specific formats

JSON

• GeoJSON

CSV

• General Transit Feed Specification (GTFS)
• Example of Prague public transport

XML

• SVG, Atom, RSS 2.0, Office Open XML (.docx, .xlsx), …

RDF

• DCAT, Schema.org

32

Generic data format properties�open vs. closed

machine-readability�binary vs. text-based

33

Open format?

34

Open vs. closed formats

Open

Specification available on the Web, freely accessible to anyone, with no limitation on its usage.

• Meta-formats e.g.: XML, JSON, CSV, RDF
• Specific formats
• SVG, GeoJSON, ...

Closed

• specification not accessible
• need for payment for access to specification
• need for registration
• need for certification of library/application claiming compatibility

Examples

• railML.org
• XML based
• need for certification

35

Machine-readable format

​

​

“All files are machine-readable, because all files are, in the end, read by machines”

36

The second part is true, but not what machine readability is about

❌

❌

❌

Machine-readable format

​

​

“Open formats like CSV, XML, JSON or RDF Turtle and Excel .xlsx files are machine readable”

37

❓

❓

❓

Machine-readable format - CSV, JSON

,,,,,,,,,,,,

Back to TOC,,,,,,,,,,,,

r2 : R2. Do you have permanent residence in Brno?,,,,,,,,,,,,

,%,count,,,,,,,,,,

Yes,89.1%,1385,,,,,,,,,,

No,10.9%,169,,,,,,,,,,

TOTAL,100.0%,1554,,,,,,,,,,

"Total sample, Weight: Weight, base n = 1554",,,,,,,,,,,,

,,,,,,,,,,,,

Back to TOC,,,,,,,,,,,,

r3 : R3. For how long have you lived in Brno? ,,,,,,,,,,,,

,%,count,,,,,,,,,,

"ico","nazev","udaje","vymazDatum","zapisDatum"

"3571092","Nadace RK CARE","[{hlavicka=Spisová značka;zapisDatum=2014-11-20;hodnotaText=N 521/KSBR;udajTyp={kod=SPIS_ZN;nazev=spisová značka};spisZn={soud={kod=KSBR;nazev=Krajský soud v Brně};oddil=N;vlozka=521}}, {hlavicka=Název;zapisDatum=2014-11-20;hodnotaText=Nadace RK CARE;udajTyp={kod=NAZEV;nazev=název}}, {hlavicka=Sídlo;zapisDatum=2014-11-20;udajTyp={kod=SIDLO;nazev=sídlo};adresa={statNazev=Česká republika;obec=Lipůvka;castObce=Lipůvka;cisloPo=385;psc=67922;okres=Blansko}}, {hlavicka=Identifikační číslo;zapisDatum=2014-11-20;hodnotaText=3571092;udajTyp={kod=ICO;nazev=identifikační číslo}}, {hlavicka=Právní forma;zapisDatum=2014-11-20;hodnotaText=nad;udajTyp={kod=PRAVNI_FORMA;nazev=právní forma};pravniForma={kod=nad;nazev=Nadace;zkratka=nad}}, {hlavicka=Účel nadace;zapisDatum=2014-11-20;udajTyp={kod=UCEL_SUBJEKTU_SEKC

38

❓

Machine-readable format - XML

<?xml version="1.0" encoding="UTF-8"?>

<PvsRejstrikData rejstrik="" operace="1" xmlns="http://portal.gov.cz/portal/xsd/PvsRejstrikData">

<TYPE>datová sada</TYPE>

<NAZEV>Smlouvy SŽDC 2017</NAZEV>

<POPIS>Uzavřené smlouvy organizace Správa železniční a dopravní cesty (resort dopravy) v roce 2017</POPIS>

<HOMEPAGE></HOMEPAGE>

<PERIODICITY></PERIODICITY>

<SPATIAL_TYPE></SPATIAL_TYPE>

<SPATIAL_TYPE_TXT></SPATIAL_TYPE_TXT>

<SPATIAL_CODE></SPATIAL_CODE>

<SPATIAL_CODE_TXT>Česká republika</SPATIAL_CODE_TXT>

<THEME></THEME>

<THEME_TXT>-</THEME_TXT>

<KEYWORDS>smlouva</KEYWORDS>

<STAV>zpracováno 2017-03-29 15:54:05</STAV>

<PROBLEMY></PROBLEMY>

<x-priloha MimeTyp="application/xml" Jmeno="data.xml">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</x-priloha>

</PvsRejstrikData>

39

❓

Machine-readable format - XLSX

40

❓

Machine-readable format

Machine readability is not a property of a format

• Depends on the form of a particular data instance
• Says whether the data is easily processed by appropriate applications
• For example
• tabular data: structured as table: rows, columns and cells well formatted, easily processed
• textual data: individual characters easily accessible, i.e. without OCR
• ...

​

41

Binary vs. text based formats

​

​

“Binary format means that the file is stored as 1s and 0s”

-- a student at a recent state exam

42

This is, of course, true also for text-based file formats

❌

❌

❌

Binary vs. text based formats

Binary files

• Their structure may be defined�on bit by bit level
• a.k.a. “non-text” file
• Not readable by text editors
• Viewable by hex editors

Text-based files

• Contains text
• Typically structured as characters on lines
• Viewable by text editors
• Also viewable by hex editors
• Text is encoded into 1s and 0s using character encoding

43

Text-based formats - character encoding - US-ASCII

Character encoding - representation of characters as binary sequences (numbers)

US-ASCII using 7 bits to represent 1 character

​

44

Original by: User:Vanessaezekowitz, CC BY-SA 3.0, via Wikimedia Commons

Text-based formats - newline representations

CR - carriage return - \r�LF - line feed - \n - Unix/Linux, MacOS�CR LF - both of them - \r\n - Windows�See all variants (Wikipedia)

45

Text-based formats - newline representations

CR - carriage return - \r�LF - line feed - \n - Unix/Linux, MacOS�CR LF - both of them - \r\n - Windows�See all variants (Wikipedia)

46

http://penguinpetes.com/

Text-based formats - character encoding - UTF-8

1 to 4 bytes representing one character

first byte compatible with US-ASCII

most frequently used characters use 2 bytes

emojis use 4 bytes

47

Character encoding - BOM - Byte order mark

Magic number at the beginning of a text file

Indicates

• unicode encoding
• encoding type
• UTF-8 - EF BB BF
• UTF-16 BE - FE FF
• UTF-16 LE - FF FE
• UTF-32 00 00 FE FF
• more...
• byte order (endianness) for multi-byte encodings

​

Most data formats use UTF-8 without BOM

• since other UTF variants are rarely used

48

Character encoding - other encodings

• ISO 646
• ASCII
• EBCDIC
• ISO 8859:
• ISO 8859-1 Western Europe
• ISO 8859-2 Western and Central Europe
• ISO 8859-3 Western Europe and South European (Turkish, Maltese plus Esperanto)
• ISO 8859-4 Western Europe and Baltic countries (Lithuania, Estonia, Latvia and Lapp)
• ISO 8859-5 Cyrillic alphabet
• ISO 8859-6 Arabic
• ISO 8859-7 Greek
• ISO 8859-8 Hebrew
• ISO 8859-9 Western Europe with amended Turkish character set
• ISO 8859-10 Western Europe with rationalised character set for Nordic languages, including complete Icelandic set
• ISO 8859-11 Thai
• ISO 8859-13 Baltic languages plus Polish
• ISO 8859-14 Celtic languages (Irish Gaelic, Scottish, Welsh)
• ISO 8859-15 Added the Euro sign and other rationalisations to ISO 8859-1
• ISO 8859-16 Central, Eastern and Southern European languages (Albanian, Bosnian, Croatian, Hungarian, Polish, Romanian, Serbian and Slovenian, but also French, German, Italian and Irish Gaelic)
• CP437, CP720, CP737, CP850, CP852, CP855, CP857, CP858, CP860, CP861, CP862, CP863, CP865, CP866, CP869, CP872
• MS-Windows character sets:
• Windows-1250 for Central European languages that use Latin script, (Polish, Czech, Slovak, Hungarian, Slovene, Serbian, Croatian, Bosnian, Romanian and Albanian)
• Windows-1251 for Cyrillic alphabets
• Windows-1252 for Western languages
• Windows-1253 for Greek
• Windows-1254 for Turkish
• Windows-1255 for Hebrew
• Windows-1256 for Arabic
• Windows-1257 for Baltic languages
• Windows-1258 for Vietnamese

​

• Mac OS Roman
• KOI8-R, KOI8-U, KOI7
• MIK
• ISCII
• TSCII
• VISCII
• JIS X 0208 is a widely deployed standard for Japanese character encoding that has several encoding forms.
• Shift JIS (Microsoft Code page 932 is a dialect of Shift_JIS)
• EUC-JP
• ISO-2022-JP
• JIS X 0213 is an extended version of JIS X 0208.
• Shift_JIS-2004
• EUC-JIS-2004
• ISO-2022-JP-2004
• Chinese Guobiao
• GB 2312
• GBK (Microsoft Code page 936)
• GB 18030
• Taiwan Big5 (a more famous variant is Microsoft Code page 950)
• Hong Kong HKSCS
• Korean
• KS X 1001 is a Korean double-byte character encoding standard
• EUC-KR
• ISO-2022-KR
• Unicode (and subsets thereof, such as the 16-bit 'Basic Multilingual Plane')
• UTF-8
• UTF-16
• UTF-32
• ANSEL or ISO/IEC 6937

49

In Czechia, from legacy systems mainly

• iso-8859-2 (Latin 2)
• windows-1250

Standardization

50

Standards - for data formats and other things

Why do we need standards?

• Interoperability, naturally
• But also business
• So it is clear who is doing something wrong...
• … and who will pay to fix it

​

51

Interoperability is costly. For each dataset:

• 1 specification created�
• each provider needs to learn the specification�
• each provider needs to adjust their data publication process�
• each consumer learns 1 specification to process all data

👨‍🔧

👨‍🔧

👩‍🔧

👨‍🔧

👩‍🔧

👩‍🔧

👩‍🔧

👨‍🔧

Spec

Low interoperability is even costlier! For each dataset:

• each provider creates specification
• each provider needs to learn the specification�
• each provider needs to adjust their data publication process�
• each consumer learns all specifications to process all data

👩‍🔧

👩‍🔧

👩‍🔧

👩‍🔧

👩‍🔧

👨‍🔧

👨‍🔧

👩‍🔧

👨‍🔧

👩‍🔧

👩‍🔧

👩‍🔧

👨‍🔧

👩‍🔧

👩‍🔧

👩‍🔧

👩‍🔧

👨‍🔧👩‍🔧

Spec

Spec

Spec

Spec

Spec

Spec

Spec

Spec

Internet Engineering Task Force - IETF

Open standards organization

• founded 1986
• initially supported by the US federal government
• now under ISOC
• participants are volunteers

IETF Working Groups

• topic, chairperson, charter, focus, deadline
• open to all

Internet Engineering Steering Group (IESG)

• final technical review of Internet standards

54

https://en.wikipedia.org/wiki/Internet_Engineering_Task_Force

Internet Society - ISOC

Americat non-profit

• founded 1992
• to provide leadership in Internet-related standards, education, access, and policy
• deals mainly with political issues
• standards are created by the �Internet Engineering Task Force (IETF) to which ISOC is related
• RFC - Request for Comments
• some become Internet Standards

“to promote the open development, evolution, and use of the Internet for the benefit of all people throughout the world”

55

https://en.wikipedia.org/wiki/Internet_Society

World Wide Web Consortium - W3C

International standards organization for the WWW

• founded 1994
• by Tim Berners-Lee - inventor of the Web
• issues Recommendations
• e.g. HTML, CSS, RDF, XML, RSS…

Specification maturation process

1. Working draft (WD)
2. Candidate recommendation (CR)
3. Proposed recommendation (PR)
4. W3C recommendation (REC)

• membership is paid and must be approved
• universities, non-profits, businesses, governments, individuals

56

https://en.wikipedia.org/wiki/World_Wide_Web_Consortium

Internet Corporation for Assigned Names and Numbers - ICANN

Standards organization

• founded 1998
• manages IANA - Internet Assigned Numbers Authority
• IPv4 and IPv6 address space management
• autonomous system number allocation
• root zone management in DNS
• …
• media types

57

https://en.wikipedia.org/wiki/ICANN

MIME-Type, Media-type

Multipurpose Internet Mail Extensions (MIME) type

The list: Media Types

Managed by�Internet Assigned Numbers Authority (IANA)

Examples:

• text/html
• text/xml
• application/xml
• application/soap+xml
• + suffix - specifies serialization - e.g. +xml, +json, +zip
• application/vnd.openxmlformats-officedocument.wordprocessingml.document
• vnd. - publicly available products, e.g. Microsoft Office
• text/x-turtle
• x- & x. - should not be used - experimental, deprecated, local, etc.

58

Ecma International

Standards organization

• founded 1961
• membership-based
• IT companies, IT trade associations, universities, foundations and public institutions
• rebranded in 1994 from�European Computer Manufacturers Association (ECMA)
• HQ: Geneva, Switzerland

Examples:

• ECMA-262 – ECMAScript Language Specification (based on JavaScript)
• ECMA-334 – C# Language Specification
• ECMA-376 – Office Open XML
• ECMA-404 – JSON

​

59

https://en.wikipedia.org/wiki/Ecma_International

RFC 2119 - Key words for use in RFCs to Indicate Requirement Levels

MUST, REQUIRED, SHALL

• an absolute requirement of the specification

MUST NOT, SHALL NOT

• an absolute prohibition of the specification

SHOULD, RECOMMENDED

• there may exist valid reasons in particular circumstances to ignore a particular item
• full implications must be understood and carefully weighed before choosing a different course

SHOULD NOT, NOT RECOMMENDED

• there may exist valid reasons in particular circumstances when the particular behavior is acceptable or even useful
• full implications should be understood and the case carefully weighed before implementing any behavior described with this label

60

RFC 5234 - Augmented Backus-Naur Form (ABNF)

Example

fragment = *( pchar / "/" / "?" )�pchar = unreserved / pct-encoded / sub-delims / ":" / "@"�pct-encoded = "%" HEXDIG HEXDIG�unreserved = ALPHA / DIGIT / "-" / "." / "_" / "~"�sub-delims = "!" / "$" / "&" / "'" / "(" / ")" / "*" / "+" / "," / ";" / "="

61

Identifiers

62

URI, URL, IRI, URN

URI - Uniform Resource Identifier - RFC 3986�URN - Uniform Resource Name - RFC 8141, IANA URN namespace registry�URL - Uniform Resource Locator - RFC 3986�IRI - Internationalized Resource Identifier - RFC 3987�� foo://example.com:8042/over/there?name=ferret#nose� \_/ \______________/\_________/ \_________/ \__/� | | | | |� scheme authority path query fragment� | _____________________|__� / \ / \� urn:example:animal:ferret:nose

63

RFC 3986 - Uniform Resource Identifier - examples

• ftp://ftp.is.co.za/rfc/rfc1808.txt
• http://www.ietf.org/rfc/rfc2396.txt
• ldap://[2001:db8::7]/c=GB?objectClass?one
• mailto:John.Doe@example.com
• news:comp.infosystems.www.servers.unix
• tel:+1-816-555-1212
• telnet://192.0.2.16:80/
• urn:oasis:names:specification:docbook:dtd:xml:4.1.2

64

RFC 3987 - IRI - Internationalized Resource Identifier

Examples

• https://opendata.gov.cz/špatná-praxe:start
• https://linked.opendata.cz/zdroj/💩
• https://en.wiktionary.org/wiki/Ῥόδος

Percent-encoding

• For some usages only URIs are acceptable
• e.g. HTTP
• IRIs are encoded in URIs
• each byte represented as �'%' and two hexadecimal digits
• e.g. 💩 => %F0%9F%92%A9
• emojis are 4 bytes in UTF-8

The same examples of IRIs percent-encoded into URIs

• https://opendata.gov.cz/%C5%A1patn%C3%A1-praxe:start
• https://linked.opendata.cz/zdroj/%F0%9F%92%A9
• https://en.wiktionary.org/wiki/%E1%BF%AC%CF%8C%CE%B4%CE%BF%CF%82

65

Tip: Copy Unicode URLs - Chrome Web Store

RFC 3492 - Punycode

IRIs not to be confused with IDN - internationalized domain name:

• e.g. https://www.háčkyčárky.cz = https://www.xn--hkyrky-ptac70bc.cz/
• even less readable than percent-encoding
• punycoded name is used with DNS

66

Data types

67

Common data types in text-based structured data formats

• boolean
• true
• false
• number
• integer
• 42
• decimal
• 42.42
• float/double
• 4.2e2

• date - ISO-8601-compliant
• YYYY-MM-DD
• 2021-03-01
• time
• HH:MM:SS.sss
• 10:40:00
• dateTime
• YYYY-MM-DDTHH:MM:SS.sss
• 2021-03-01T10:40:00
• time zones
• 2021-03-01T10:40:00+02:00
• 2021-03-01-02:00
• 2021-03-01Z

The same data types used in all common formats - RDF syntaxes, XML, JSON, CSV�Based on XML Schema data type system

68