Class: XData::FileReader
Constant Summary collapse
- RE_Y =
/lat|(y.*coord)|(y.*pos.*)|(y.*loc(atie|ation)?)/i
- RE_X =
/lon|lng|(x.*coord)|(x.*pos.*)|(x.*loc(atie|ation)?)/i
- RE_GEO =
/^((geom.*)|location|locatie|coords|coordinates)$/i
- RE_NAME =
/(title|titel|naam|name)/i
- RE_A_NAME =
/^(naam|name|title|titel)$/i
- GEOMETRIES =
["point", "multipoint", "linestring", "multilinestring", "polygon", "multipolygon"]
Instance Attribute Summary collapse
-
#content ⇒ Object
readonly
Returns the value of attribute content.
-
#file ⇒ Object
readonly
Returns the value of attribute file.
-
#params ⇒ Object
readonly
Returns the value of attribute params.
Instance Method Summary collapse
- #download ⇒ Object
- #fillOut ⇒ Object
- #find_col_sep(f) ⇒ Object
- #find_geometry(xfield = nil, yfield = nil) ⇒ Object
- #find_unique_field ⇒ Object
- #findExtends ⇒ Object
- #geom_from_text(coords) ⇒ Object
- #get_address ⇒ Object
- #get_fields ⇒ Object
- #guess_name ⇒ Object
- #guess_srid ⇒ Object
-
#initialize(pars) ⇒ FileReader
constructor
A new instance of FileReader.
- #is_geo_json?(s) ⇒ Boolean
- #is_wkb_geometry?(s) ⇒ Boolean
- #is_wkt_geometry?(s) ⇒ Boolean
- #odata_json(url) ⇒ Object
- #parseODataFields(props) ⇒ Object
- #parseODataMeta(md) ⇒ Object
- #proces_zipped_dir(d) ⇒ Object
- #read_csv(path, c = nil) ⇒ Object
- #read_json(path, hash = nil) ⇒ Object
- #read_odata(h) ⇒ Object
- #read_shapefile(path) ⇒ Object
- #read_xdata(path) ⇒ Object
- #read_xml(path, data = nil) ⇒ Object
- #read_zip(path, data = nil) ⇒ Object
- #srid_from_prj(str) ⇒ Object
- #write(path = nil) ⇒ Object
Constructor Details
#initialize(pars) ⇒ FileReader
Returns a new instance of FileReader.
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 |
# File 'lib/xdata/file_reader.rb', line 85 def initialize(pars) @params = pars if @params[:file_path] =~ /^http(s)?:\/\/.+/ download else file_path = File.(@params[:file_path]) if File.extname(file_path) == '.xdata' read_xdata(file_path) else ext = @params[:originalfile] ? File.extname(@params[:originalfile]) : File.extname(file_path) case ext when /\.zip/i read_zip(file_path) when /\.(geo)?json/i read_json(file_path) when /\.shp/i read_shapefile(file_path) when /\.csv|tsv/i read_csv(file_path) when /\.xdata/i read_xdata(file_path) when /\.xml/i read_xml(file_path) else raise "Unknown or unsupported file type: #{ext}." end end end fillOut end |
Instance Attribute Details
#content ⇒ Object (readonly)
Returns the value of attribute content.
32 33 34 |
# File 'lib/xdata/file_reader.rb', line 32 def content @content end |
#file ⇒ Object (readonly)
Returns the value of attribute file.
32 33 34 |
# File 'lib/xdata/file_reader.rb', line 32 def file @file end |
#params ⇒ Object (readonly)
Returns the value of attribute params.
32 33 34 |
# File 'lib/xdata/file_reader.rb', line 32 def params @params end |
Instance Method Details
#download ⇒ Object
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 |
# File 'lib/xdata/file_reader.rb', line 55 def download() data = '' if @params[:file_path] =~ /\/ODataFeed\// open(odata_json(@params[:file_path])) do |f| data = XData::parse_json(f.read) read_json(nil,data) end else open(@params[:file_path]) do |f| data = f.read end if @params[:file_path] =~ /\.csv$/i read_csv(nil,data) elsif @params[:file_path] =~ /\.zip$/i read_zip(nil,data) elsif data =~ /^\s*<\?xml/ read_xml(nil,data) else begin data = XData::parse_json(data) read_json(nil,data) rescue XData::Exception return end end end fillOut end |
#fillOut ⇒ Object
34 35 36 37 38 39 40 41 42 |
# File 'lib/xdata/file_reader.rb', line 34 def fillOut @params[:rowcount] = @content.length get_fields unless @params[:fields] guess_name unless @params[:name] guess_srid unless @params[:srid] find_unique_field unless @params[:unique_id] get_address unless @params[:hasaddress] findExtends unless @params[:bounds] end |
#find_col_sep(f) ⇒ Object
197 198 199 200 201 202 203 204 |
# File 'lib/xdata/file_reader.rb', line 197 def find_col_sep(f) a = f.gets b = f.gets [";","\t","|"].each do |s| return s if (a.split(s).length == b.split(s).length) and b.split(s).length > 1 end ',' end |
#find_geometry(xfield = nil, yfield = nil) ⇒ Object
294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 |
# File 'lib/xdata/file_reader.rb', line 294 def find_geometry(xfield=nil, yfield=nil) delete_column = (@params[:keep_geom] != true) return if @content.blank? unless(xfield and yfield) @params[:hasgeometry] = nil xs = true ys = true @content[0][:properties].each do |k,v| next if k.nil? if k.to_s =~ RE_GEO srid,g_type = is_wkb_geometry?(v) if(srid) @params[:srid] = srid @params[:geometry_type] = g_type @content.each do |h| a,b,g = is_wkb_geometry?(h[:properties][k]) h[:geometry] = g h[:properties].delete(k) if delete_column end @params[:hasgeometry] = k return true end srid,g_type = is_wkt_geometry?(v) if(srid) @params[:srid] = srid @params[:geometry_type] = g_type @content.each do |h| a,b,g = is_wkt_geometry?(h[:properties][k]) h[:geometry] = g h[:properties].delete(k) if delete_column end @params[:hasgeometry] = k return true end srid,g_type = is_geo_json?(v) if(srid) @params[:srid] = srid @params[:geometry_type] = g_type @content.each do |h| h[:geometry] = h[:properties][k] h[:properties].delete(k) if delete_column end @params[:hasgeometry] = k return true end end hdc = k.to_s.downcase if hdc == 'longitude' or hdc == 'lon' or hdc == 'x' xfield=k; xs=false end if hdc == 'latitude' or hdc == 'lat' or hdc == 'y' yfield=k; ys=false end xfield = k if xs and (hdc =~ RE_X) yfield = k if ys and (hdc =~ RE_Y) end end if xfield and yfield and (xfield != yfield) @params[:hasgeometry] = [xfield,yfield] @content.each do |h| h[:properties][xfield] = h[:properties][xfield] || '' h[:properties][yfield] = h[:properties][yfield] || '' h[:geometry] = {:type => 'Point', :coordinates => [h[:properties][xfield].gsub(',','.').to_f, h[:properties][yfield].gsub(',','.').to_f]} h[:properties].delete(yfield) if delete_column h[:properties].delete(xfield) if delete_column end @params[:geometry_type] = 'Point' @params[:fields].delete(xfield) if @params[:fields] and delete_column @params[:fields].delete(yfield) if @params[:fields] and delete_column return true elsif (xfield and yfield) # factory = ::RGeo::Cartesian.preferred_factory() @params[:hasgeometry] = [xfield] @content.each do |h| h[:geometry] = geom_from_text(h[:properties][xfield]) h[:properties].delete(xfield) if h[:geometry] and delete_column end @params[:geometry_type] = '' @params[:fields].delete(xfield) if @params[:fields] and delete_column return true end false end |
#find_unique_field ⇒ Object
134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 |
# File 'lib/xdata/file_reader.rb', line 134 def find_unique_field fields = {} @params[:unique_id] = nil @content.each do |h| h[:properties].each do |k,v| fields[k] = Hash.new(0) if fields[k].nil? fields[k][v] += 1 end end fields.each_key do |k| if fields[k].length == @params[:rowcount] @params[:unique_id] = k break end end end |
#findExtends ⇒ Object
270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 |
# File 'lib/xdata/file_reader.rb', line 270 def findExtends geometries = [] if @params[:hasgeometry] @content.each do |o| o[:geometry][:type] = 'MultiPolygon' if o[:geometry][:type] == 'Multipolygon' geometries << Geometry.from_geojson(o[:geometry].to_json) end geom = GeometryCollection.from_geometries(geometries, (@params[:srid] || '4326')) @params[:bounds] = XData.toPolygon(geom.bounding_box()) elsif @params[:postcode] pc = @params[:postcode].to_sym @content.each do |o| p2 = PC4.lookup(o[:properties][pc]) if p2 geometries << GeoRuby::SimpleFeatures::Point.from_coordinates(p2[0], (@params[:srid] || '4326')) geometries << GeoRuby::SimpleFeatures::Point.from_coordinates(p2[1], (@params[:srid] || '4326')) end end geom = GeometryCollection.from_geometries(geometries, (@params[:srid] || '4326')) @params[:bounds] = XData.toPolygon(geom.bounding_box()) end end |
#geom_from_text(coords) ⇒ Object
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 |
# File 'lib/xdata/file_reader.rb', line 251 def geom_from_text(coords) # begin # a = factory.parse_wkt(coords) # rescue # end if coords =~ /^(\w+)(.+)/ if GEOMETRIES.include?($1.downcase) type = $1.capitalize coor = $2.gsub('(','[').gsub(')',']') coor = coor.gsub(/([-+]?[0-9]*\.?[0-9]+)\s+([-+]?[0-9]*\.?[0-9]+)/) { "[#{$1},#{$2}]" } coor = JSON.parse(coor) return { :type => type, :coordinates => coor } end end {} end |
#get_address ⇒ Object
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 |
# File 'lib/xdata/file_reader.rb', line 117 def get_address pd = pc = hn = ad = false @params[:housenumber] = nil @params[:hasaddress] = 'unknown' @params[:postcode] = nil @params[:fields].reverse.each do |f| pc = f if ( f.to_s =~ /^(post|zip|postal)code.*/i ) hn = f if ( f.to_s =~ /huisnummer|housenumber|(house|huis)(nr|no)|number/i) ad = f if ( f.to_s =~ /address|street|straat|adres/i) end if pc and (ad or hn) @params[:hasaddress] = 'certain' end @params[:postcode] = pc @params[:housenumber] = hn ? hn : ad end |
#get_fields ⇒ Object
166 167 168 169 170 171 172 173 174 175 |
# File 'lib/xdata/file_reader.rb', line 166 def get_fields @params[:fields] = [] @params[:alternate_fields] = {} return if @content.blank? @content[0][:properties].each_key do |k| k = (k.to_sym rescue k) || k @params[:fields] << k @params[:alternate_fields][k] = k end end |
#guess_name ⇒ Object
153 154 155 156 157 158 159 160 161 162 163 164 |
# File 'lib/xdata/file_reader.rb', line 153 def guess_name @params[:name] = nil @params[:fields].reverse.each do |k| if(k.to_s =~ RE_A_NAME) @params[:name] = k return end if(k.to_s =~ RE_NAME) @params[:name] = k end end end |
#guess_srid ⇒ Object
177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 |
# File 'lib/xdata/file_reader.rb', line 177 def guess_srid return if @content.blank? return unless @content[0][:geometry] and @content[0][:geometry].class == Hash @params[:srid] = 4326 g = @content[0][:geometry][:coordinates] if(g) while g[0].is_a?(Array) g = g[0] end lon = g[0] lat = g[1] if lon.between?(-7000.0,300000.0) and lat.between?(289000.0,629000.0) # Simple minded check for Dutch new rd system @params[:srid] = 28992 end else end end |
#is_geo_json?(s) ⇒ Boolean
231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 |
# File 'lib/xdata/file_reader.rb', line 231 def is_geo_json?(s) return nil if s.class != Hash begin if GEOMETRIES.include?(s[:type].downcase) srid = 4326 if s[:crs] and s[:crs][:properties] if s[:crs][:type] == 'OGC' urn = s[:crs][:properties][:urn].split(':') srid = urn.last.to_i if (urn[4] == 'EPSG') elsif s[:crs][:type] == 'EPSG' srid = s[:crs][:properties][:code] end end return srid,s[:type],s end rescue Exception=>e end nil end |
#is_wkb_geometry?(s) ⇒ Boolean
206 207 208 209 210 211 212 213 214 215 216 |
# File 'lib/xdata/file_reader.rb', line 206 def is_wkb_geometry?(s) begin f = GeoRuby::SimpleFeatures::GeometryFactory::new p = GeoRuby::SimpleFeatures::HexEWKBParser.new(f) p.parse(s) g = f.geometry return g.srid,g.as_json[:type],g rescue => e end nil end |
#is_wkt_geometry?(s) ⇒ Boolean
218 219 220 221 222 223 224 225 226 227 228 |
# File 'lib/xdata/file_reader.rb', line 218 def is_wkt_geometry?(s) begin f = GeoRuby::SimpleFeatures::GeometryFactory::new p = GeoRuby::SimpleFeatures::EWKTParser.new(f) p.parse(s) g = f.geometry return g.srid,g.as_json[:type],g rescue => e end nil end |
#odata_json(url) ⇒ Object
44 45 46 47 48 49 50 51 52 53 |
# File 'lib/xdata/file_reader.rb', line 44 def odata_json(url) if url =~ /\/ODataFeed\// uri = URI.parse(url) return url + '?$format=json' if uri.query.nil? pars = CGI.parse(uri.query) return url if pars["$format"] return url + '&$format=json' end return url end |
#parseODataFields(props) ⇒ Object
494 495 496 497 498 499 500 501 502 503 |
# File 'lib/xdata/file_reader.rb', line 494 def parseODataFields(props) rank=1 @params[:md] = {} if @params[:md].nil? props.each do |p| @params[:md]["fieldUnit.#{rank}".to_sym] = p[:Unit] @params[:md]["fieldDescription.#{rank}".to_sym] = p[:Description] @params[:md]["fieldLabel.#{rank}".to_sym] = p[:Key] rank += 1 end end |
#parseODataMeta(md) ⇒ Object
481 482 483 484 485 486 487 488 489 490 491 492 |
# File 'lib/xdata/file_reader.rb', line 481 def parseODataMeta(md) @params[:md] = {} if @params[:md].nil? @params[:md][:title] = md[:Title] @params[:md][:identifier] = md[:Identifier] @params[:md][:description] = md[:Description] @params[:md][:abstract] = md[:ShortDescription] @params[:md][:modified] = md[:Modified] @params[:md][:temporal] = md[:Period] @params[:md][:publisher] = md[:Source] @params[:md][:accrualPeriodicity] = md[:Frequency] @params[:md][:language] = md[:Language] end |
#proces_zipped_dir(d) ⇒ Object
615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 |
# File 'lib/xdata/file_reader.rb', line 615 def proces_zipped_dir(d) Dir.foreach(d) do |f| next if f =~ /^\./ if File.directory?(d + '/' + f) return true if proces_zipped_dir(d + '/' + f) end case File.extname(f) when /\.(geo)?json/i read_json(d+'/'+f) return true when /\.shp/i read_shapefile(d+'/'+f) return true when /\.csv|tsv/i read_csv(d+'/'+f) return true end end return false end |
#read_csv(path, c = nil) ⇒ Object
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 |
# File 'lib/xdata/file_reader.rb', line 385 def read_csv(path, c = nil) if path File.open(path, "r:bom|utf-8") do |fd| c = fd.read end end unless @params[:utf8_fixed] detect = CharlockHolmes::EncodingDetector.detect(c) c = CharlockHolmes::Converter.convert(c, detect[:encoding], 'UTF-8') if detect end c = c.force_encoding('utf-8') c = c.gsub(/\r\n?/, "\n") @content = [] @params[:format] = 'CSV' @params[:colsep] = find_col_sep(StringIO.new(c)) unless @params[:colsep] csv = CSV.new(c, :col_sep => @params[:colsep], :headers => true, :skip_blanks =>true) csv.header_convert { |h| h.blank? ? '_' : h.strip.gsub(/\s+/,'_') } csv.convert { |h| h ? h.strip : '' } index = 0 begin csv.each do |row| r = row.to_hash h = {} r.each do |k,v| h[(k.to_sym rescue k) || k] = v end @content << {properties: h } index += 1 end rescue => e raise XData::Exception.new("Read CSV; line #{index}; #{e.message}") end find_geometry end |
#read_json(path, hash = nil) ⇒ Object
421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 |
# File 'lib/xdata/file_reader.rb', line 421 def read_json(path, hash=nil) STDERR.puts hash.class if hash @content = [] if path data = '' File.open(path, "r:bom|utf-8") do |fd| data = fd.read end hash = XData::parse_json(data) end if hash.is_a?(Hash) and hash[:'odata.metadata'] read_odata(hash) elsif hash.is_a?(Hash) and hash[:type] and (hash[:type] == 'FeatureCollection') # GeoJSON hash[:features].each do |f| f.delete(:type) @content << f end @params[:hasgeometry] = @params[:format] = 'GeoJSON' else # Free-form JSON @params[:format] = 'JSON' val,length = nil,0 if hash.is_a?(Array) # one big array val,length = hash,hash.length else hash.each do |k,v| if v.is_a?(Array) # the longest array value in the Object val,length = v,v.length if v.length > length end end end if val val.each do |h| @content << { :properties => h } end end find_geometry end end |
#read_odata(h) ⇒ Object
505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 |
# File 'lib/xdata/file_reader.rb', line 505 def read_odata(h) @content = [] @params[:format] = "OData" @params[:odata] = {} links = h[:value] links.each do |l| @params[:odata][l[:name].to_sym] = l[:url] end begin open(odata_json(@params[:odata][:TableInfos])) do |f| md = XData::parse_json(f.read)[:value] parseODataMeta(md[0]) end open(odata_json(@params[:odata][:DataProperties])) do |f| props = XData::parse_json(f.read)[:value] parseODataFields(props) end open(odata_json(@params[:odata][:TypedDataSet])) do |f| c = XData::parse_json(f.read)[:value] c.each do |h| @content << { :properties => h } end end rescue OpenURI::HTTPError => e STDERR.puts e. end find_geometry end |
#read_shapefile(path) ⇒ Object
540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 |
# File 'lib/xdata/file_reader.rb', line 540 def read_shapefile(path) @content = [] prj = path.gsub(/.shp$/i,"") + '.prj' prj = File.exists?(prj) ? File.read(prj) : nil srid_from_prj(prj) if (prj and @params[:srid].nil?) @params[:hasgeometry] = 'ESRI Shape' @params[:format] = "Shape File" GeoRuby::Shp4r::ShpFile.open(path) do |shp| shp.each do |shape| h = {} h[:geometry] = XData::parse_json(shape.geometry.to_json) #a GeoRuby SimpleFeature h[:properties] = {} att_data = shape.data #a Hash shp.fields.each do |field| s = att_data[field.name] s = s.force_encoding('ISO8859-1') if s.class == String h[:properties][field.name.to_sym] = s end @content << h end end end |
#read_xdata(path) ⇒ Object
608 609 610 611 612 |
# File 'lib/xdata/file_reader.rb', line 608 def read_xdata(path) h = Marshal.load(File.read(path)) @params = h[:config] @content = h[:content] end |
#read_xml(path, data = nil) ⇒ Object
567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 |
# File 'lib/xdata/file_reader.rb', line 567 def read_xml(path, data=nil) if path File.open(path, "r:bom|utf-8") do |fd| data = fd.read end end begin feed = Feedjira::Feed.parse(data) if feed @params[:format] = "Atom Feed" maxlat = -1000 maxlon = -1000 minlat = 1000 minlon = 1000 doc = Nokogiri::XML data a = doc.xpath("//georss:polygon") if a.length > 0 # geometries << GeoRuby::SimpleFeatures::Point..from_latlong(lat, lon) a.each do |x| # 50.6 3.1 50.6 7.3 53.7 7.3 53.7 3.1 50.6 3.1 s = x.text.split(/\s+/) s.each_slice(2) { |c| maxlat = [maxlat,c[0].to_f].max maxlon = [maxlon,c[1].to_f].max minlat = [minlat,c[0].to_f].min minlon = [minlon,c[1].to_f].min } end @params[:bounds] = { type: 'Polygon', coordinates: [[minlon,minlat], [minlon,maxlat], [maxlon,maxlat], [maxlon,minlat], [minlon,minlat]] } end end # url = feed.entries[0].url # Dir.mktmpdir("xdfi_#{File.basename(path).gsub(/\A/,'')}") do |dir| # f = dir + '/' + File.basename(path) # end rescue Exception => e puts e.inspect return -1 end end |
#read_zip(path, data = nil) ⇒ Object
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 |
# File 'lib/xdata/file_reader.rb', line 639 def read_zip(path, data=nil) tempfile = nil begin if(data) tempfile = Tempfile.new('xdatazip') tempfile.write(data) path = tempfile.path end Dir.mktmpdir("xdfi_#{File.basename(path).gsub(/\A/,'')}") do |dir| command = "unzip '#{path}' -d '#{dir}' > /dev/null 2>&1" raise XData::Exception.new("Error unzipping #{path}.", {:originalfile => path}, __FILE__, __LINE__) if not system command if File.directory?(dir + '/' + File.basename(path).chomp(File.extname(path))) dir = dir + '/' + File.basename(path).chomp(File.extname(path) ) end return if proces_zipped_dir(dir) end rescue Exception => e raise XData::Exception.new(e., {:originalfile => path}, __FILE__, __LINE__) ensure tempfile.unlink if tempfile end raise XData::Exception.new("Could not process file #{path}", {:originalfile => path}, __FILE__, __LINE__) end |
#srid_from_prj(str) ⇒ Object
469 470 471 472 473 474 475 476 477 478 479 |
# File 'lib/xdata/file_reader.rb', line 469 def srid_from_prj(str) begin connection = Faraday.new :url => "http://prj2epsg.org" resp = connection.get('/search.json', {:mode => 'wkt', :terms => str}) if resp.status.between?(200, 299) resp = XData::parse_json resp.body @params[:srid] = resp[:codes][0][:code].to_i end rescue end end |
#write(path = nil) ⇒ Object
665 666 667 668 669 670 671 672 673 674 675 676 |
# File 'lib/xdata/file_reader.rb', line 665 def write(path=nil) path = @file_path if path.nil? path = path + '.xdata' begin File.open(path,"w") do |fd| fd.write( Marshal.dump({:config=>@params, :content=>@content}) ) end rescue return nil end return path end |